Browny
Smash Hater
I've been wanting to write a theory guide for a while now and I've finally though of a topic to talk about and that's how to get better at reads and punishes, something I think that a lot of players starting to become competitive think that just happens naturally over time. While that is true and it does, you can actively work to improve your punishes by learning what the opponent can do.
As always with my threads, its theorycraft ahead. This is the sort of theorycraft that is backed by my regularly playing the best players in my state offline and online as I have been doing for many, many years so if you don't like theorycraft, that's fine. I don't like it either most of the time. That said, this is how I view things and no one has to agree with it, but I find it works very well for me and if there's one thing I'm ever complimented on my gameplay, its how I can read people that I've never played before. Hopefully new and developing competitive players might find a use for it.
Firstly the obvious way to land a punish on your enemy is by reading their habits and predicting an option. It is not that easy to pull off in practice but you can make this step far easier and it begins with understanding the steps before that. In order to punish the enemy for doing anything you need to understand the following:
1) Know your enemies options
2) Reduce them
3) Read them
4) Punish them
Once you know this, you practice it and develop a solid reaction time which is made easier through step 2 and you will find you can land hard punishes from reads more often and be a much more difficult player to beat.
Let just get right into it.
I believe that many people focus too hard on what their character can do at the expense of learning other characters and this directly results in you being at a disadvantage when attempting to read the enemy. If you know your characters combos like muscle memory, you can do all of the advanced techs, you watch all the top videos of that character and you are in complete control of your movement, you are only in control of 50% of the variables in a match. The other 50% is the enemy. They are in control, they know how to combo you, they know the techs and most importantly, they know how to move to avoid you.
If the enemy knows what you can do, but you don't know what they can do, they are going to have a significant advantage which is simply called matchup experience. This however is completely separate to reading and punishing the enemy. Someone can play against a certain character for hours on end every day but if they don't actively attempt to read and punish them, they are going to let the enemy get away far too often with weak or missed punishes. Further to this someone else playing the character can have a completely unique and different playstyle which can throw off your experience. You shouldn't focus on what the enemy is doing in any given situation, instead focus on what they CAN do and what you can do about it It requires dedication and focus to get reads regularly, it doesn't come for free. You need to be actively thinking about it during the entire match.
The first step to getting better at reads and punishes is knowing your enemies options at all times. For many people the idea of leaning how to play each character to a high level is utterly infeasible so you must rely on playing other people who use the character or watching videos of the matchup. What is absolutely critical though is if for example in my case, I am watching a video of a Mewtwo vs Falco, I spend an equal amount of time watching what the Falco does, than watching the Mewtwo. Even better, watch the entire match once from the perspective of your main, and then re-watch it from the perspective of the enemy.
You want to watch how they approach, how they defend. What options THEY choose. You will notice that in certain situations, some options are better than others and they are more regularly chosen. The situations we are looking at, in order to land punishes, are disadvantage states for the enemy. That is the state where they have been hit away, are near the edge or just generally on the back foot from your pressure. This entire essay is on that state. Lets look at the situation of Falco being at the disadvantage state of being off-stage however he still has his double jump and sideb to recover.
You may watch or practice this scenario and think to yourself “If the enemy chooses option A, I am going to react with Punish A. If they choose B, punish with option B”. With falco off-stage with a double jump, he has about 5 options with jump then phantasm, phantasm to the ledge, recover low with phantasm-jump, recover low with phantasm into upb or jump-upb to the ledge. These sheer amount of options mean you should not be attempting to simply predict and react in order to punish because it is just too difficult to do, running a 20% guess where choosing the wrong option can get you spiked with his sideb is not a good scenario to be in, you need it to be better than that. I'm going to leave this at that but come back to it later.
Playing a reactionary game in order to land reads and punishes is going to result in you missing a lot of opportunities because so many characters have so many mix-ups that reacting is a futile exercise, the success rate will be low. When you aren't making reactions, you are making reads with result in harder punishes and you do this by step 1, reducing the enemies options.
When the enemy is in their disadvantage state you should no longer focus on what your character can do and instead focus on what the enemy can do. You should quickly be able to switch your point of view to the enemy and think from their point of view. Lets look at examples as we go along. Dr Mario is on the edge of the stage and you are DK spamming dtilt on his shield. The immediate thing that should come into your mind is “If I keep doing this, he has to do something to get out. If I were Dr Mario, I know that I can upb out of shield, roll behind DK, hold shield or spotdodge. Nothing else will work” You then also know that your spacing is correct, upb out-of-shield wont connect and his roll dodge doesn't have enough range to get behind DK. As the dtilts keep coming and his shield gets smaller, Dr Mario is running out of time and now his only option is to spotdodge and that is when you read it with the hardest punish, an fsmash.
This situation is complex. Attempting to read an early spotdodge with an fsmash is a risky move because Dr Mario has the option of continuing to hold shield and that would result in you getting punished. By knowing all of Dr Mario's options, the DK player knows that continuing to use dtilt over and over despite it hitting his shield is going to cut down Dr Mario's options down from 4, to 2, to 1. When the enemy has literally no option left, you do not need to rely on a reaction to secure a hard punish, you have reduced their options so far that you can easily read a spotdodge because they cant do anything else giving you the punish of your choice.
There are countless situations this can apply in. In most situations you are not going to be able to cut down an enemies options to one unless a ledge or bad recovery is involved but you should always strive to cut down their options as much as possible. The most usual situation this applies to is juggling the enemy. When the enemy is trying to land they are extremely vulnerable to dashing upsmashes and projectiles.
Lets look at the case of Ike on FD at kill %'s after being knocked upwards out of his jump and you are Ness. Ike knows that there is a grab or a upair coming his way. As soon as he is up you should understand his options. Your first mission is to quickly determine his options; Ike can land with nothing, fair, nair, sideb away, counter, aether or eruption. So he has 7 options which he can use to rather easily beat your attempted kill move. Your next mission is to cut off as many of those as possible, but how do you do it?
By running behind Ike, you have cut off aether, sideb and fair while giving him bair. Still that is 7 options down to 5. You now have the choice of going for a uair kill or waiting for him to land. But think of Ike's options. Nair, counter and eruption beat a uair attempt, but bair beats a grab attempt. This is where you fake them out with a short hop. By jumping upwards, the Ike knows that he has lost the bair option as he cant hit you with it unless he is spacing it as he lands, the only reason you did this was to take away that option, forcing Ike to rely on a prediction. He loses that but gains airdodge to dodge an upair however that will result in him getting heavy landing lag. Now Ike has 3 options in counter, nair and eruption. Instead, you simply land back on the ground with shield. This removes all of Ike's safe options immediately and you have an easy grab on Ike as he lands. The read and punish is complete whether he dodges or attacks because your option beats all of them and you achieved because you knew their options.
In less than a second, you cut the enemies options from from 7 to 5, then to 3 which are all punishable by the same move. Through spacing alone the enemy often loses many options and if you exert any pressure on them whatsoever you can generally cut down another option. At no time have you relied on a mere prediction to get a 1/7 guess right, you guaranteed the punish and to the viewer it almost looks like the Ike player gave you the punish for free by falling for something so obvious. In reality though, you successfully knew their options, reduced them, read them and punished.
I can go on all day with examples but I think the point I am trying to make is done. The key to everything here is knowing the enemies options. Only when you know their options can you cut them down. Remember that you have no more than 50% influence in how any match is done and only focusing on your character opens you up to getting read like a book while you have no reads of your own, only lucky predictions.
In many instances though actively reducing options down to 1 will give you a larger amount of reads. Even in situations where the enemy may be cut down to 2 options this is where you rely on conditioning the opponent by watching their particular habits and going for hard reads. A hard read is basically a read where the enemy had a variety of options but you correctly predicted the right one and had a punish ready. There is a difference here. No matter who is behind the controller, a character always has the same options. In situations where an enemy has 2 or more viable options though this is where player habits come into play giving you a greater chance in landing a hard read.
When you become very adept at cutting down the enemies options, you will naturally become better at reading them. To maximise the effectiveness of this you need to be aware of the strongest punishes. This doesn't need detailed knowledge of the enemy though, this part is all about your character. When you are confident in your ability to force and react to an enemies options, go for hard punishes. Your success rate of landing a hard punish can quickly rise from 20 to >50% by understanding what the enemy can do.
I hope that is clear to people now and they can approach learning the game in a more advanced way. You can get better at reading and punishing if you put the effort in to learn how characters work and shift your focus from your perspective, to their perspective and react accordingly. They key to all of this obviously is learning the enemies options. I truly believe the best way to do this is to watch videos of high level players using characters and try to emulate their style. Even when they are beating up on your main, imagine the game from their perspective. Pause videos and rewind to think about what they could have done different to avoid getting punished. The sorts of things you wouldn't want your opponent to know. Critically, you don't need regular matchup experience versus a character in order to be fully aware of their options. This is a very... very big point I am trying to make here.
Just before I finish, the earlier example of Falco off stage vs Mewtwo. How might you approach this? He has five options (jump then phantasm, phantasm to the ledge, recover low with phantasm-jump, recover low with phantasm into upb or jump-upb to the ledge.), and you want the hardest punish. This is how I would do it.
As always with my threads, its theorycraft ahead. This is the sort of theorycraft that is backed by my regularly playing the best players in my state offline and online as I have been doing for many, many years so if you don't like theorycraft, that's fine. I don't like it either most of the time. That said, this is how I view things and no one has to agree with it, but I find it works very well for me and if there's one thing I'm ever complimented on my gameplay, its how I can read people that I've never played before. Hopefully new and developing competitive players might find a use for it.
Firstly the obvious way to land a punish on your enemy is by reading their habits and predicting an option. It is not that easy to pull off in practice but you can make this step far easier and it begins with understanding the steps before that. In order to punish the enemy for doing anything you need to understand the following:
1) Know your enemies options
2) Reduce them
3) Read them
4) Punish them
Once you know this, you practice it and develop a solid reaction time which is made easier through step 2 and you will find you can land hard punishes from reads more often and be a much more difficult player to beat.
Let just get right into it.
I believe that many people focus too hard on what their character can do at the expense of learning other characters and this directly results in you being at a disadvantage when attempting to read the enemy. If you know your characters combos like muscle memory, you can do all of the advanced techs, you watch all the top videos of that character and you are in complete control of your movement, you are only in control of 50% of the variables in a match. The other 50% is the enemy. They are in control, they know how to combo you, they know the techs and most importantly, they know how to move to avoid you.
If the enemy knows what you can do, but you don't know what they can do, they are going to have a significant advantage which is simply called matchup experience. This however is completely separate to reading and punishing the enemy. Someone can play against a certain character for hours on end every day but if they don't actively attempt to read and punish them, they are going to let the enemy get away far too often with weak or missed punishes. Further to this someone else playing the character can have a completely unique and different playstyle which can throw off your experience. You shouldn't focus on what the enemy is doing in any given situation, instead focus on what they CAN do and what you can do about it It requires dedication and focus to get reads regularly, it doesn't come for free. You need to be actively thinking about it during the entire match.
The first step to getting better at reads and punishes is knowing your enemies options at all times. For many people the idea of leaning how to play each character to a high level is utterly infeasible so you must rely on playing other people who use the character or watching videos of the matchup. What is absolutely critical though is if for example in my case, I am watching a video of a Mewtwo vs Falco, I spend an equal amount of time watching what the Falco does, than watching the Mewtwo. Even better, watch the entire match once from the perspective of your main, and then re-watch it from the perspective of the enemy.
You want to watch how they approach, how they defend. What options THEY choose. You will notice that in certain situations, some options are better than others and they are more regularly chosen. The situations we are looking at, in order to land punishes, are disadvantage states for the enemy. That is the state where they have been hit away, are near the edge or just generally on the back foot from your pressure. This entire essay is on that state. Lets look at the situation of Falco being at the disadvantage state of being off-stage however he still has his double jump and sideb to recover.
You may watch or practice this scenario and think to yourself “If the enemy chooses option A, I am going to react with Punish A. If they choose B, punish with option B”. With falco off-stage with a double jump, he has about 5 options with jump then phantasm, phantasm to the ledge, recover low with phantasm-jump, recover low with phantasm into upb or jump-upb to the ledge. These sheer amount of options mean you should not be attempting to simply predict and react in order to punish because it is just too difficult to do, running a 20% guess where choosing the wrong option can get you spiked with his sideb is not a good scenario to be in, you need it to be better than that. I'm going to leave this at that but come back to it later.
Playing a reactionary game in order to land reads and punishes is going to result in you missing a lot of opportunities because so many characters have so many mix-ups that reacting is a futile exercise, the success rate will be low. When you aren't making reactions, you are making reads with result in harder punishes and you do this by step 1, reducing the enemies options.
When the enemy is in their disadvantage state you should no longer focus on what your character can do and instead focus on what the enemy can do. You should quickly be able to switch your point of view to the enemy and think from their point of view. Lets look at examples as we go along. Dr Mario is on the edge of the stage and you are DK spamming dtilt on his shield. The immediate thing that should come into your mind is “If I keep doing this, he has to do something to get out. If I were Dr Mario, I know that I can upb out of shield, roll behind DK, hold shield or spotdodge. Nothing else will work” You then also know that your spacing is correct, upb out-of-shield wont connect and his roll dodge doesn't have enough range to get behind DK. As the dtilts keep coming and his shield gets smaller, Dr Mario is running out of time and now his only option is to spotdodge and that is when you read it with the hardest punish, an fsmash.
This situation is complex. Attempting to read an early spotdodge with an fsmash is a risky move because Dr Mario has the option of continuing to hold shield and that would result in you getting punished. By knowing all of Dr Mario's options, the DK player knows that continuing to use dtilt over and over despite it hitting his shield is going to cut down Dr Mario's options down from 4, to 2, to 1. When the enemy has literally no option left, you do not need to rely on a reaction to secure a hard punish, you have reduced their options so far that you can easily read a spotdodge because they cant do anything else giving you the punish of your choice.
There are countless situations this can apply in. In most situations you are not going to be able to cut down an enemies options to one unless a ledge or bad recovery is involved but you should always strive to cut down their options as much as possible. The most usual situation this applies to is juggling the enemy. When the enemy is trying to land they are extremely vulnerable to dashing upsmashes and projectiles.
Lets look at the case of Ike on FD at kill %'s after being knocked upwards out of his jump and you are Ness. Ike knows that there is a grab or a upair coming his way. As soon as he is up you should understand his options. Your first mission is to quickly determine his options; Ike can land with nothing, fair, nair, sideb away, counter, aether or eruption. So he has 7 options which he can use to rather easily beat your attempted kill move. Your next mission is to cut off as many of those as possible, but how do you do it?
By running behind Ike, you have cut off aether, sideb and fair while giving him bair. Still that is 7 options down to 5. You now have the choice of going for a uair kill or waiting for him to land. But think of Ike's options. Nair, counter and eruption beat a uair attempt, but bair beats a grab attempt. This is where you fake them out with a short hop. By jumping upwards, the Ike knows that he has lost the bair option as he cant hit you with it unless he is spacing it as he lands, the only reason you did this was to take away that option, forcing Ike to rely on a prediction. He loses that but gains airdodge to dodge an upair however that will result in him getting heavy landing lag. Now Ike has 3 options in counter, nair and eruption. Instead, you simply land back on the ground with shield. This removes all of Ike's safe options immediately and you have an easy grab on Ike as he lands. The read and punish is complete whether he dodges or attacks because your option beats all of them and you achieved because you knew their options.
In less than a second, you cut the enemies options from from 7 to 5, then to 3 which are all punishable by the same move. Through spacing alone the enemy often loses many options and if you exert any pressure on them whatsoever you can generally cut down another option. At no time have you relied on a mere prediction to get a 1/7 guess right, you guaranteed the punish and to the viewer it almost looks like the Ike player gave you the punish for free by falling for something so obvious. In reality though, you successfully knew their options, reduced them, read them and punished.
I can go on all day with examples but I think the point I am trying to make is done. The key to everything here is knowing the enemies options. Only when you know their options can you cut them down. Remember that you have no more than 50% influence in how any match is done and only focusing on your character opens you up to getting read like a book while you have no reads of your own, only lucky predictions.
In many instances though actively reducing options down to 1 will give you a larger amount of reads. Even in situations where the enemy may be cut down to 2 options this is where you rely on conditioning the opponent by watching their particular habits and going for hard reads. A hard read is basically a read where the enemy had a variety of options but you correctly predicted the right one and had a punish ready. There is a difference here. No matter who is behind the controller, a character always has the same options. In situations where an enemy has 2 or more viable options though this is where player habits come into play giving you a greater chance in landing a hard read.
When you become very adept at cutting down the enemies options, you will naturally become better at reading them. To maximise the effectiveness of this you need to be aware of the strongest punishes. This doesn't need detailed knowledge of the enemy though, this part is all about your character. When you are confident in your ability to force and react to an enemies options, go for hard punishes. Your success rate of landing a hard punish can quickly rise from 20 to >50% by understanding what the enemy can do.
I hope that is clear to people now and they can approach learning the game in a more advanced way. You can get better at reading and punishing if you put the effort in to learn how characters work and shift your focus from your perspective, to their perspective and react accordingly. They key to all of this obviously is learning the enemies options. I truly believe the best way to do this is to watch videos of high level players using characters and try to emulate their style. Even when they are beating up on your main, imagine the game from their perspective. Pause videos and rewind to think about what they could have done different to avoid getting punished. The sorts of things you wouldn't want your opponent to know. Critically, you don't need regular matchup experience versus a character in order to be fully aware of their options. This is a very... very big point I am trying to make here.
Just before I finish, the earlier example of Falco off stage vs Mewtwo. How might you approach this? He has five options (jump then phantasm, phantasm to the ledge, recover low with phantasm-jump, recover low with phantasm into upb or jump-upb to the ledge.), and you want the hardest punish. This is how I would do it.
Throw an uncharged shadow ball. This cuts down the option of phantasm to the ledge instantly which is his safest recovery. This opens up 2 paths;
If falco jumps, throw another uncharged shadow ball. This covers him falling before going for a ledgesnap. He does not want to be clipped by another one offstage as he wont recover, he is going to go for a phastasm above your head. His options have gone from 5, to 2, to 1. Read and punish this option with an upsmash.
If falco goes for a low recovery, throw another one since he can simply jump then phantasm. His options are now phantasm-jump or phantasm upb. Fall offstage with a nair to clip him out of his jump, KO'ing him.
Such a perfect theoretical punish is not exactly easy to pull off in practice because you will rarely be at the perfect spacing to pull off these punishes but the point is you have maximised your chance of getting the right read and at no time have put yourself in a position where you risk getting spiked by his phantasm. You have achieved the strongest punish with the lowest amount of risk achieved by knowing their options with no predictions necessary.
If falco jumps, throw another uncharged shadow ball. This covers him falling before going for a ledgesnap. He does not want to be clipped by another one offstage as he wont recover, he is going to go for a phastasm above your head. His options have gone from 5, to 2, to 1. Read and punish this option with an upsmash.
If falco goes for a low recovery, throw another one since he can simply jump then phantasm. His options are now phantasm-jump or phantasm upb. Fall offstage with a nair to clip him out of his jump, KO'ing him.
Such a perfect theoretical punish is not exactly easy to pull off in practice because you will rarely be at the perfect spacing to pull off these punishes but the point is you have maximised your chance of getting the right read and at no time have put yourself in a position where you risk getting spiked by his phantasm. You have achieved the strongest punish with the lowest amount of risk achieved by knowing their options with no predictions necessary.
Last edited:
