Counterintuitive Things in Melee

Practicing something for hours on end does not make you perfect at it. There are several reasons for this. When practicing something for extensive periods of time, you get in a groove that you will never be in at a tournament. Moreover, stimulus conditions vary greatly. You will never be in that same mental state at a tournament. You will be focused on playing your opponent, not on ledgedashing. Doing things in tournament is different than in your room. You aren’t going to get nervous and choke in your room. It’s best to practice things frequently in short bursts, for ~10 minutes a day, or 10 minutes a few different times in a day.

Copying the best players doesn’t make you one of them. In other games, you can fake it until you make it. It can be ok to just emulate top players and let the understanding follow as your execution improves. This does not work well in Melee. Even if you think you are copying top players, what they are doing is probably different from what you are doing. They have extensive understanding of situations and their opponent’s mental state. If you try to do what Norwalk players do, you are probably just going to look stupid. If you are still bent on copying top players, pick ones that play standard, optimal styles, e.g. Wizzrobe.

Striking to bad stages can be good. If your opponent has 2 counterpicks in a BO3 and you are confident he will pick the one you don’t ban, consider starting on that stage. Often, the way you play on bad counterpicks doesn’t say much about your overall strategy. You might learn your opponent better there than he can learn you. You also might prefer losing game 1 and having a winnable game 3 over having to play on their best CP game 3. Similar logic can apply to BO5. You can still opt to strike to something more neutral in any case but don’t neglect the possibility of striking to a bad stage if you are going to have to play on it later in the set anyway.

Longer dashes aren’t always better. Having a character that can crouch out of run quickly offers them options that characters with longer dashes don’t have. Cactuar dash offers a similar movement option to dash dance for these characters. It lets you implement crouching into your movement very nicely. Even if your character has a long dash, be sure to work on dash dancing different lengths, because sometimes a good tight dash dance can be indispensable.

Sometimes it’s better to not hit someone; sometimes getting hit is good. In doubles, hitting someone out of the corner with a move that doesn’t hit them far off the other side of the stage can be really bad. Sometimes hitting someone lets their team regain stage position. In singles, sometimes hitting people ends up with you in a worse position as well. You can go off with a weak knee, then let your opponent recover before you. Sometimes hitting people lets them DI and slide to the ledge, so it might have been better to do something like let their shield decay more instead of going for the poke. Not all openings are equal. Be calculated about how you want to hit someone. Crouch canceling really comes into play here, because sometimes the person who lands the first hit can immediately be punished, even if they try to do it safely.

Guaranteed followups can be worse than not guaranteed ones. Opportunities are fleeting in Melee. Sometimes you have reads and you can be confident that you are correct. In these situations, going for what’s guaranteed can be squandering an opportunity that you might not get again. If you always go for a regrab, you are giving your opponents extra chances to DI to safety. Seize opportunities when they are there, and end your opponent’s stocks when given the chance.

Not approaching can be good even when you’re losing. Melee has 8 minutes on the clock. Taking 1-2 minutes off by forcing the game to be played at your pace can be extremely helpful. By slowing the game down, you can make the opponent play at your pace, not theirs. Most people get frustrated and approach. Unless your opponent is actually going to wait out the clock (I’m one of the only people in NorCal who is disciplined enough to not get cheesed by people like Laudandus who try to abuse this), you can get away with camping when you’re losing.

Doing things that “should never work” can be super good. Top players, especially ones from WC like Shroomed, S2J, Mango, and Lord do all kinds of things that “shouldn’t” work. But they read their opponent, sensing either their habits in game, or their mental state outside of the game, and know they can get away with things. You can even condition your opponents to expect something standard and sensible, then flip the situation on its head by doing something really stupid that they don’t expect. This works really well if you can sense that your opponent is nervous or choking. You can actually read that someone is choking or that they are going to flub an input. Sometimes doing something that “shouldn’t work” leads to the best possible outcome. If you are unwilling to do things because they shouldn’t work against some TAS bot, you are squandering opportunites. No Melee player is perfect. Look for the imperfections of your opponent and take advantage (but don’t do this habitually when playing friendlies against noobs because then you’ll develop numerous bad habits).

Sometimes doing moves extremely slowly is good. Delaying moves to the last possible frames often looks stupid, but if you get good at it, the rewards will be huge. People have timing windows where they expect their opponents to throw out their moves. This is only natural because things like dash dances and jumps have a rhythm to them. Play outside the conventional rhythm and you can make people look stupid . They end up getting caught or baited by unexpected timings. Even something as stupid as missing the dash dance or “Gould dashing” can lead to a timing mixup, where something that looks retarded ends up getting rewarded (RIP Kalamazhu If you are skeptical about this point or the previous one, just remember that Mattdotzeb beat Leffen.

Not all kills are equal. Emotional damage is a thing. You can cause people to crumble by ending their stocks in painful ways. The mental game is a thing, and you can gain mental advantages by devastating opponents in unexpected ways. Have you ever seen someone SD and then fall apart right after that? Or how about that time Darkrain phantomed his kill move, then was unable to regain any footing for the rest of the match? It’s possible to intentionally cause these decisive moments in the match. If you see a way to get a particularly humiliating kill, or a chance to kill your opponent in a way he may have never seen before, go for it. Getting into your opponent’s head is a serious advantage.

Key differences from singles to keep in mind when playing teams

There is no way to be a remotely good singles player without an excellent understanding of neutral game. Neutral game is possibly the most important part and probably the most difficult part (playing against someone else is harder than playing against yourself). If you are playing a support role in doubles, it’s possible to win playing no neutral game whatsoever. Singles is about a dance to jockey for advantageous positioning. You play patient and find your spot, then take your punish as far as it goes. Doubles is about trying to create situations where your opponents have no options whatsoever. Besides wobbling and a few chaingrabs, you can pretty much always do something to escape situations in singles. In doubles, there are countless situations where there is absolutely nothing you can do. If you are playing against a team that isn’t on point about capitalizing on overextensions to create their own “100-0” situation, it’s worth playing risky and aggressive. Most teams are not on point about turning reversals into no win situations. Situational understanding is probably the most important thing in singles, but this is a subject where there is little consensus. Situational understanding is by far and away, unequivocally the most important thing to playing doubles.

In singles, there are objectively and obvious best situations. Doubles has a lot of situations where you have alternative options that are all good, and it’s not clear what to do. Learning these ahead of times. Because you have a partner to back you up, the risk/reward of certain options is completely different from in singles. Because the logic of which options are good in singles doesn’t necessarily make sense in the context of doubles, you really need to figure out things ahead of time. Coordination is paramount, so you and your partner need to be on the same page about which options you favor in which situations. Doubles is much more about improvisation and much more free flow than singles. This may seem to contradict what I just said, but it doesn’t. Singles lets you get away with camping into flowchart punishes if you are good at staying safe and have good execution. Simply because there are innumerable situations with 4 players on the screen compared to 2, doubles forces countless novel situations, which you need to learn to navigate.

I’ve mentioned this in every guide, but time is the most important resource in doubles. Earning time to look at the entire screen in order to assess where to position yourself is a victory in itself. Learn which situations allow you to take advantage of the time you’ve earned, and don’t squander advantages you’ve gained by choosing options that take too much time. Waiting is bad. Doubles favors being proactive far more than singles. If you don’t feel the urgency when you are playing, odds are that you have poor situational awareness. There’s almost always useful things to be done on the screen. How to watch the screen while watching or playing doubles is a skill to be learned. There are focal points that your eyes should be drawn to if you understand what’s going on. Check out some of my doubles content on to get a better idea of what to look at on the screen during teams.

How to learn the neutral game

I have recently realized that people are not proficient at using their brain while playing. Things that seem obvious or self evident to some are not intuitive to others. Thus, I’m writing a short article about how to think about Melee. As the title suggests, this is mainly regarding neutral game, but also has relevance with ledge guards and punishes in general.

If you talk to people about Melee and what the most important parts of the required skillset are, you’ll here vastly different opinions. Some of the best players are unable to articulate even basic things about how they play and why they do what they do. But if you don’t feel like you are playing with an intuitive understanding of what’s going on, you are going to need to exercise your critical thinking skills.

I believe the most important skill to be a high level Melee player is understanding situations. Second to that, and first in the minds of many players, is movement. People know to grind movement, but without any knowledge of how to apply your movement, you probably won’t be able to get ranked in any remotely competitive region. Dr. PP once pointed out that many players are really good at looking cool between stocks but have no idea how to apply it in their play. Don’t let this be you.

Lord is the player I’m going to talk about here, though I may analyze other Falcon players in future articles. The guy has great tech skill, but it’s nothing crazy and he doesn’t do much next level stuff in terms of movement. In reality, he probably moves less than any other top Falcon ever has. The only thing that he really has over other Falcons is choosing the right choices in microsituations (which goes hand in hand with his excellent use of crouch cancel). The guy is sometimes able to win sets against top players just by reaching deep into his bag of gimmicks and repeatedly tricking them in a few situations he has a better understanding of. He spams all kinds of unsafe things that can be countered, but he is good at reading you and knows janky options that beat all the conventional ones.

The point of this is that you should be continually thinking about what your opponent is doing and how to counter it. If you aren’t picking options that specifically counter things that you think your opponent is going to do, you are either being wishful that your opponent is bad or you are hax$. When you get hit, ask yourself why. Try to come up with answers for the situation in which you got hit, or try to figure out how to avoid the situation altogether. Don’t be afraid to solicit advice from better players. Melee nerds love talking about Melee.

Ultimately, fighting games are all games of weighted rock paper scissors. Certain options beat certain options but the risk/reward is skewed. Don’t be afraid to throw something out that’s high risk if you have a strong reason to believe it’s going to beat what they are going to do. It’s very rare to find someone whose actions are totally unpredictable. Look for situations where you can take advantage of the habits of the opponent, and look to avoid doing any defensive things because of autopilot. If I ask you “what does that option beat?” you should have an answer related to what you have been seeing your opponent doing. It should be obvious how this relates to ledgeguarding, which is frequently one of the areas that could use the most improvement in the skillsets of Melee players.

Improving Efficiency of Practice by NMW

Coming from an RTS background and being good at a lot of games, I’ve always had a lot of thoughts about how to get good at Melee. I wrote this guide some time ago, and though I still believe this methodology was good for me, I am more cognizant than ever that it is not right for everyone. One of the main deterrents for improvement in Melee is the quality of practice partners, and my old guide takes that into account. I now have significant experience in tutoring people at Melee, and education in general, which have given me more thoughts about how to improve. Tools for practicing have also come a long way since I wrote that article.

People suck at getting good at games. There are innumerable reasons. This guide is written with the assumption that you have regular access to some practice partners that aren’t complete noobs and that any individual reading it does not lack proper motivation. Mentality problems are a common reason for stifled improvement that I won’t discuss here. Coming from RTS, I’m used to hearing people should blindly copy top players until they understand them, i.e. fake it till you make it by emulating good players. This is awful advice in Melee. I would know because I tried to emulate SilentSpectre when I started. This was among the worst ideas I’ve had in my Melee career. Doing things that you don’t understand the purpose behind is terrible in Melee. Your hands should never move faster than your brain, unless you’re one of the 1% of people like n0ne or Mango who are actually good at playing solely on feel.

Practice useful things. W33dl0rd sits in my living room all day practicing the most useless shit. It can be difficult as a newer player to recognize what’s useful and what isn’t. I too spent a ton of my time practicing useless shit. If you aren’t at least an upper-mid level player, it’s unlikely that any of your ideas for next level tech are useful. If you are struggling to find useful things to practice, look at the paragons of your character—Leffen, Wizzrobe, Plup, etc. Though no one has a maxed out toolkit, watching any of these optimal/standard players should give you plenty to practice. If you have practiced something to the point that you can execute it, but after a month you are still unable to integrate it into your gameplay, it might be time to abandon that idea. Not every technique needs to fit into your style. You’ll probably learn some stuff during attempted implementation even if you are unsuccessful. Ledge play is something most people with most characters could spend a lot of time polishing. Check out Laudandus’s practice regiment from a few months ago to see examples of useful things

If you are a newcomer, the single best thing you could do to improve is to get 20xx and to turn on the color overlay for wait state and missed L-cancels. Get rid of all your idle frames. The color overlay does not work when you’re airborne or in IASA frames, so be sure to not neglect your fastfalls or movement out of grounded moves like Marth’s dtilt. Despite the fact that you can improve at Melee in infinitely many ways, this is the thing that will make the most difference out the gate. The two most important things for being good at Melee are movement and knowledge of situations. The first you can grind out and rapidly improve at, and the second comes through experience and having an active mind, both in the game and between sessions.

Playing Melee is not necessarily practicing Melee. Your brain should be fully engaged and you should have several things in mind you want to practice going into every session. Your skill level and your partner’s skill level should directly influence this. If you are still new, you should practice tech skill, DI, and punish game. At any level, there’s still punish mixups and tech skill to work on. The worse your opponent is, the more you should focus on these things. If you are a mid level player against another mid level player or a better player, neutral game will probably occupy a large portion of your cognitive load. Improving your neutral game is something that is harder to dedicate focus to within a practice regiment, and an area where you will probably do much of your learning implicitly. If you are just playing fuck around, not very serious games, don’t dedicate more than half of your focus to neutral. Neutral game is best practiced when at peak focus against better players who are actively trying to counter you. If you aren’t really in the zone, it’s probably best to focus on practicing punish game and other technical techniques because these are the areas where practice most readily translates into glaring improvements.

Even if your punish game is good, it might not be good against someone with better defense than your common opponents. Remember that punish and offensive mixups are likely to work on bad players but may not work on better ones. Thus, work on recognizing the situations, and work on the execution, but don’t get it worked into your muscle memory to indiscriminately cover certain scrubby things, like instant double jumps or rolls towards center. Better players navigate defensive situations much better than your nooby friends. Good examples of situations where the disparity in defensive skills are the most salient are recovering when off stage, getting out of the corner, and moving oos (especially after a late aerial on shield). Always be diligent in trying to think about how someone might escape and what options might cover their escape. Your friends may not have good SDI, but top players do, so if you’re trying to get better, you need to develop ideas for how to cover options against better players, even if you don’t come across these options often.

Part of getting the most out of your practice is recognizing which practice on your opponent will generalize to other players. The fewer practice partners you have, the more you will develop metagames just to beat them, which may even harm you when you play someone else. For instance, I live with Shroomed, who likes to run at you and fight. I might be very good against him, but I could be very bad against Sheik in general. If I am only exposed to a style where my opponent runs at me, I am likely to have no idea how to deal with planking, platform camping, standing in the corner and throwing needles, etc. When people say “x is so good at a matchup because they practice with y,” this is almost always bullshit because they are just good at the player matchup. No one plays every style of a a character. Overstating the relevance of character matchups and understating the value of style vs. style or player vs. player matchups is among the most common fallacies in our scene. People always stupidly talk up their friends for being good at a matchup, then they come across styles they haven’t played before, and they look like they have no idea wtf they are doing, e.g.

In order to figure out what is useful to practice against which opponents, categorize things you want to practice. The order of categories here correlate to the skill level of your opponent, i.e. focus entirely on the first categories if your opponents are bad, and work your way up as the quality of your opponents increases.

  1. Things that are useful against everyone, like spacing moves at their tip, grabbing the ledge quickly, and ledgedashing.
  2. Guaranteed punishes, bearing in mind that things that feel guaranteed against your friends may not actually be guaranteed.
  3. Things that are generally good.  You want to avoid spamming gimmicks because that will inevitably stifle your improvement. There is a big difference between things that work and things that are good. Good is mostly a function of risk reward, so good things should be hard to punish, or they should have an extremely high reward based on you having a correct and not outlandish read. Basically, you want to cover as many options as safely as possible, or you want your option to cover an option with extreme potency.
  4. Ways to trick your opponent. Actual mindgames and trickery are unlikely to be relevant until a higher level.  At lower levels you can just wait for someone to do something stupid, and even if they don’t do stupid things, you still might struggle to trick someone because they lack situational understanding. Someone might actually be too stupid to roll or spotdodge because they don’t understand they’re pressured!
  5. If playing against someone who’s much better than you, put extensive focus on getting out of bad situations by doing things like good SDI, crisp WDs OOS, Amsah techs, etc.  This builds situational awareness.

Not included in these categories is experimentation of next level shit. Do that when you are feeling yourself and have moments of inspiration. These are likely to come out when you are noobstomping, but sometimes it can happen against better players. Or maybe you are just drunk and everything feels like it’s working magically. Don’t be hesitant to record yourself if you aren’t in a sober state. It’s not unlikely that this is the time when you learn the most from watching yourself play.

Contact NMW on Twitter @NMWhittier or @NMWFalcon, and subscribe on YouTube for content.

Tournament Preparation Guide

HugS, Tafokints, and others have written guides to tournament preparation. While some may find them helpful, I know they would not be helpful for me, so I am guessing I’m not alone in this demographic. Existing guides mostly talk about prudent planning, like eating right and wearing deodorant. While I don’t necessarily disagree with what’s been written in the past, I can tell you that none of these guides would have helped me improve at the game. Thus, I want to share some tips both for how to play your best in tournament and how to get the most out of attending a tournament.

How to do your best in this particular tournament:

Prepare in a way that makes meaningful differences for you. Likely factors here are sleep, practice, drugs, and food. I happen to know that practice is infinitely more important for me than sleep or food. I might even play my best on no food or sleep but with binge practice as a substitute. Experiment as you would scientifically by controlling variables. I can’t play at all hungover, but there probably are some people who can. I’m not saying you should do anything that will adversely affect your health, but you should try to make an effort to learn what hinders you. There’s a reason why Scar used to spend half of his time on commentary whining about being hungry. If you go to Melee tournaments often, you will constantly be put into situations that could make you have a bad performance, so you want to learn to navigate these situations.

Warmup time tends to be difficult to get, so be assertive about getting on setups. I play best when I have played as much as possible with as few breaks as possible. I wish this were not the case, but I need to have my hands constantly warm to play my best. Don’t lose because you were too pussy to ask to get next. Look for setups that will get you the practice you need. Find the matchups that you want to practice or the setup with the opponents that look like they will best prepare you for coming hours. This might be the setup with the best players so you learn the most, or it might be the setup with the worst players so you get your hands moving their fastest. Don’t focus too far ahead on the bracket though. If you are thinking about the Samus you have to beat in WR3, you are getting taken out of the zone and might get upset in WR1 or WR2. There’s nothing shittier than letting yourself get upset because you assumed you couldn’t lose.

Develop a routine to help you get in the zone. This may involve isolating yourself from friends for a little bit to focus. Maybe you want to review your notes. Maybe you want to get baked and combo the worst players there. There’s no right answer here. Just experiment a lot and see what helps you play your best. There are endless possibilities for how to get into the zone. Just remember that standing around annoyed or bored makes it difficult to transition to good Melee. We can’t all play our best all the time, but you can get to know yourself so that you have the best chance to play your best. Johns are real and they make people play bad, but no one cares, so do your best to put yourself in a situation where you don’t have anything to John about. Don’t be the guy who whines about the venue being cold—you could have brought a sweater and hand warmers!

Practice things in game that help you perform better. Use your handwarmers, morning routine, or whatever time you have to practice the things that will make you win. This might sound obvious, but it’s a little less straightforward than it seems. Haxdashes, for instance, are a fairly insignificant tool in my kit. However, being unable to perform them under pressure is something that breaks my mental game. If I can’t perform my tech skill under pressure, my focus breaks and my game falls apart. Make sure you can perform everything you expect yourself to be able to perform. On top of this, practice things that are truly game winning, like waveshine into usmash, just to make sure you don’t throw games. Making sure you are on top of your basics is the best way to get ready for the impending mental battle. In your match, you want to focus solely on your opponent and you don’t want to have to spend any focus on execution.

People generally play their best Melee when they are having fun. People play badly when they are nervous or taken out of the zone by external factors. If you are thinking about what people are going to say about the result of the match, you are not playing your best. If you are thinking about a combo that you want to do to look super sick, you are not playing your best. Hopefully this guide can help you find the fun so that you will be able to play your best more often.

Don’t say to yourself “I will win” or “I need to win.” Think “I’ll do my best,” “I’m prepared,” remind yourself of strategies, etc. “Will” and “need” are sentiments that take you out of the game. “I’ll do my best” might not be the best thing for you think, but it works pretty well for me because it doesn’t take me out of the game.  Thoughts about the outcome of the game do.  You should be solely focused on the movements of your opponent and on his mental state.  Focus is the most important factor in winning. Winning isn’t about being better, it’s about taking four stocks faster than they take four of yours. Being taken out of the game mentally leaves you susceptible to missing something (simple shit like an SDI input can win a game). Conversely, if your opponent’s focus breaks, this is an opportunity for you. Experiment playing with music and maybe even game sound in headphones to see if these things help you focus. Even if you rely on audio cues, you will be a more formidable adversary if you are well-focused and lacking audio cues than if you had audio cues but poor focus.

How to set yourself up in the long term:

Let’s start with some basic tournament etiquette, for the newcomers or the socially daft. Be calculated with when you talk to top players, i.e. don’t talk to them after they’ve lost or while they are preparing for an important match. When you do talk to them, don’t say anything too fanboyish and don’t come off too strong. Remember, these are normal guys who weren’t so different from you when you started, so it’s more likely you’ll be their friends if you just act like a normal fucking person. If you come off as starstruck, you might make them less inclined to hang out because you’ve made it awkward for them. Also, be wary of calling next when everyone on the setup is vastly more skilled than you. You might be impinging on their much needed preparation time, though you shouldn’t be too shy to ask to money match or see that this isn’t the case.

Be sparing with your shit talk. It’s fine to shit talk your friends if you’re funny and having a good time, but don’t be a dick to strangers. If you’re shit talking your friend’s opponent in r1 losers, you might actually be making a newcomer’s tournament experience shitty. Melee hasn’t gotten where it is now by being mean to noobs. Help cultivate the scene by being a decent person.

Let people money match on the setup if you’re playing friendlies. There’s a basic hierarchy of importance that tournament > mms > friendlies. If you want to use the setup, money match yourself. Don’t do 1 dollar BO5; that is the grimiest shit. If you want to play a couple games against a friend you haven’t seen in years, or something extenuating like that, people will understand, but generally be ready to vacate the setup if two people want to MM.

Expect winner stays on rotation. Melee is a fighting game. Fighting games have decades of history of winner staying on the rotation. If someone wins several rounds of the rotation and you have a desire to play someone else on the rotation, speak up. Most people don’t have a problem sitting out a game if they win ten in a row. If you really want to play someone, money match.

Ask questions when you play. Be specific. Examples of good questions include “What other options do I have in that situation?” or “do you notice any glaring bad habits that I have?” Examples of bad questions include “What do I do against lasers” or “How do I beat Fox?” Don’t argue with top players when they try to give you advice—it’s usually better to be quiet and let the advice stir, as you may come to understand it as time passes. If you come off as particularly scrubby as a result of having poorly thought out questions, the person is probably going to walk off and tell his friends about the retarded scrub he was playing. People LIKE motivated youngins. If you come off like you have the fire and a good attitude, better players are more likely to play you regardless of the skill disparity.

If all the setups look occupied and you like doubles, walk around and suggest to people to play doubles until you find a setup that agrees. If you have one person winning a setup who wants to play doubles, you only need you, your partner, and then one more random person to convert that setup to doubles. Because doubles usually comes first, people are more likely to be inclined to play doubles before the tournament starts, which is often the time with the most friendlies. Just getting your hands warm will benefit you more than doing nothing, even if you aren’t entering doubles. If you are a new player, your singles skills will increase more by playing doubles continuously than they would losing every game on a four man rotation. Playing doubles can also be a fun way to bond, get other people more into the game, and to bring people out of their social shell.

Be social. If you want to get good, it’s in your best interest to mingle because this gets you practice partners. This is doubly true if you are a new player and trying to get into a clique of top players. Top players don’t hang out with each other because they think they are so cool and so good at the game. It’s because they’ve been coming to the same events for years and know each other quite well. If you’re a newcomer, you are unlikely to offer them much of anything in the game, so building rapport outside the game is the best way to posture to get the proper practice partners for long term improvement. This is probably the most important point in this guide, so let me drive it in with an anecdote. A few months ago, I spent an hour at a setup with W33dl0rd, an acquaintance who is not very good, and a noob I had never met. The noob took 0 stocks and he asked 0 questions, just being awkward, not laughing with the rest of us when we were shooting the shit or interacting with us in any way. He was about to take one stock, then he taunted, missed the kill, and proceeded to get four stocked. He probably learned nothing from sitting there on a setup with three better players for an hour. He made himself look like a complete fucking idiot with that taunt, showed no signs whatsoever of being cool, and I certainly did not learn anything from playing him. Why the fuck would I ever invite this guy to a smashfest?

Every tournament has an ideal way to enjoy it. Some tournaments will involve lots of friendlies, some involve lots of money matches, Evo involves excessive drinking, etc. Discuss in advance on social media which tournaments are worth going to if you’re inexperienced. If you are flying to Evo to play lots of friendlies, you are making a mistake. If you are going to The Foundry to practice, you are making a mistake. When in Rome, do as the Romans do. Every tournament is imperfect. Develop your groove to have the most fun at tournaments you frequent. Find good food places, figure out what the schedule is actually like, figure out which people you want to practice with go to which tournaments, etc. If I go to Big House this year, I plan on playing at least fifty money matches. I know that’s how I will get the most out of that tournament.

Contact NMW on Twitter @NMWhittier or @NMWFalcon, and subscribe on YouTube for content.

NMW’s Advanced Guide to Doubles

This is the third guide in my series for learning doubles.  If you haven’t read the first two, I highly recommend doing so, as they will be more useful in terms of how much you will get out of it for amount of time spent reading.  See Teams 101: and Teams 102:  Here are some basic reminders from previous teams guides: time is your most important resource so don’t waste it, don’t get yourself cornered going for ledgeguards during invincibility, let yourself reset to neutral while letting your teammate start offense should you find yourself at a neutral game disadvantage. On that last note, don’t do this. These two consecutive defensive choices by SFAT are literally game losing:  I honestly believe that doubles is easier than singles but requires much more knowledge.  If you understand the situations in this guide, you will be an amazing doubles player.

Maximizing Damage Dealt:

Since time is the most important resource in doubles, you have to learn how to use your time most efficiently. In game, this means using moves that do the most knockback and percent, and that are most likely to lead to quick kills. Team UGS is the best team to study in terms of efficient punishes, e.g. PewFat is good as well: That said, don’t get too caught up in the flashiness of team UGS or PewFat. Many times, their opponents do a poor job of disrupting their combos. In actuality, you will often have to let a combo go sooner than you see in these GIFs, so it’s often fine to just use your finisher as soon as possible to earn stage positioning and additional time to double team their teammate.

No situation exemplifies efficient time use better than the seconds after a whiffed rest. Here’s what not to do: In this situation, the Puff was at a high enough percent so a fully charged Falcon fsmash would kill. However, a Fox usmash would kill with minimal charge. Thus, the best thing to do would be for red team to attack the Blue Falcon. This is an uninterruptable 2v1. After maybe 2 interactions, the Red Fox should then go to usmash the Puff. This gives two potential kills and makes one relatively guaranteed (look at how far the Blue Falcon was initially). Instead, knowing that Red Falcon wants to get a kill, Blue Falcon is emboldened to go for an extremely risky knee, knowing the Red Falcon is tunnel visioned on the prospect of a low % kill on a Puff. Here’s another one where I am able to read the hunger of my opponent to kill the sleeping Puff and turn the situation around: Conversely, these are examples of perfect rest punishes:

A meta part of using time efficiently is learning when you have the chance to eye the entire screen. Many situations earn you enough time to eye the screen. Here are situations where there is no justifiable reason not to eye the entire screen: Getting hit far off stage, grabbing the ledge, landing a grab, landing a launcher, shielding, jumping to a platform. Other times, you should eye the entire screen, but a lapse is forgivable. These situations include landing a tech chase, landing a waveshine, landing a multihit move, hitting an opponent off stage, and getting hit. Remember that a lot of the time, it’s favorable to keep getting hit if your partner is fucking their partner up faster than you are getting fucked up. Conversely, in massively advantageous offensive positions, you should still make sure your partner isn’t going to die faster than you are going to do damage. In my observations, the most common wastes of time are trying to pressure people with ledge invincibility and waiting below opponents trying to shark their landings (for the love of god, don’t do this). Here’s efficient time use: Hit targets of opportunity. It’s usually better to hit who’s near you rather than chasing someone across the stage. Fighting close to your teammate facilitates teamwork and maximum brutality.

One of the most efficient possible uses of time is the sandwich combo. Here is the sandwich combo in its purest form: The Blue Fox takes 64 damage in < 1 second. That means he now has to be worried about getting killed already. Don’t position yourself like the Fox did there—you should avoid sandwiches because they are the best way to get fast damage. Here’s another GIF exemplifying how devastating sandwich combos can be: A conventional way of starting a sandwich would be to snipe the support, hitting him away, and then double teaming the remaining player between you and your teammate. In this clip, Shroomed tries to maximally push an advantage, trying to create a sandwich on Armada while positioning to ledge guard Android, but Armada knows that Shroomed loves to run past the front line, so he’s ready right away and punishes Shroomed for his bold positioning And here’s one more showing PewPewU being crafty and creating a sandwich situation: In this situation, not only does PPU create a sandwich by going for the back line, but he throws one so that both opponents are in the sandwich, creating an opportunity to get massive damage.

Advanced spacing with your teammate:

There’s several ways to space around your teammate. The most important way is covered in the optimal positioning section in Teams 101. Staying about one Marth tipper range (slightly closer even) behind your teammate if you’re the support is an ideal position to react. It lets you collapse on them if they gain space or prevent loss of space should they lose the neutral. However, once you are at an advantage, there’s three main ways to space. One is to space just outside your teammate’s hitboxes, as seen in the gif with MacD and Vanz. These are the most basic sandwich combos.

Sandwiches have a number of subtleties to them. Move selection immediately comes to mind. Hitboxes, hurtboxes, knockback distance, and knockback trajectory are all important concerns to keep in mind when choosing your moves. Big hitboxes are a double edged sword because, while they make it easier to hit the opponent, they also require more precise spacing to avoid hitting your teammate. Hurtboxes are important because certain moves change your hurtbox and your teammate will probably not be ready to space around this. Disjoints are obviously great for avoiding this, but take Marth’s fsmash for example. Though it’s as disjointed as any other move in the game, after the hitbox has ended, Marth still leans forward, greatly changing his hurtbox. Try not to use moves that leave your hurtbox extended. As for knockback, you need to be careful to not use moves with too much knockback at medium percents, or else you will hit your opponents out of the sandwich.

The second way of spacing is the way Teams UGS does it, spacing at the edge of the knockback of your teammate’s moves, which is important for maintaining sandwiches with slower characters. Positioning during sandwiches is slightly trickier than it may seem. Basic sandwiches where you space at the edge of your teammate’s attacks and spam moves are easy. But in order to get the team UGS combos, you will have to move away from your teammate. Practice recognizing when you should position for passes instead of just spamming moves next to your teammate. Team UGS always position themselves at the ideal distance. Peach is slow, so this is especially important when teaming with her, though this is made easier by the fact that Peach and Sheik don’t have great disparities in knockback within their movesets. As far as trajectory goes, moves with the typical spacie bair trajectory are ideal for sandwiches. They hit people up, but not too far up. This is important because horizontal knockback is preferable over vertical in most doubles situations, but moves that hit with downwards trajectories like Sheik’s fair may allow your opponent to tech.

The third way to space offensively is the way Javi and Twin do it, and isn’t too different from the first, except they space around moving bodies, which is immensely more difficult. It’s difficult to articulate, so just check out some of their gifs. This really requires acting as one mind, but it has massive rewards: These guys have no fear. Though it’s not possible for me to give you a perfect guide on how to replicate their spacing, I can give a few rules of thumb:

  1. Whenever you are in front and you have the opportunity to space on the far side of the opponent, do it. This makes it possible for the back line to follow and space behind you. If they SDI behind, your teammate will be there. It also makes it so that, if you hit with a move like Falco dair, they are hit towards your teammate and not away.
  1. Designate roles. It makes it easier to be one mind if you know what to expect. If you know one person is taking lead, it makes it easier to anticipate when they will swing. If you know they are trying to swing and that they will crossup when possible, this makes it easier to collapse and space with them.
  2. Drift early in your moves, not late. If you telegraph your spacings at startup, your teammate can space confidently with you in a no risk way. When you feel there’s no risk, you can go for stylish things without second guessing yourself.

Similar principles apply to shield pressure. If a Fox hits a shield, his teammate should be able to confidently collapse on the Fox, and position to cover a roll, but also be ready to start sandwich shield pressure if there looks like there’s an opportunity to shield poke. Crossing up when you’re in front is a powerful way to include your teammate in offense. If you are in the back and know your teammate will not cross up, sometimes it is ok to jump over them and start sandwich pressure that way, though ideally you would do this for a shield stab, not just to hit a shield, because you are forfeiting your roll coverage. Moves with large shield stun, such as Falcon’s knee, can be used more liberally in this way, because if you stagger shield stun correctly with your teammate’s attacks, you can make it very hard for the opponent to escape.

The last tenet of team spacing is remembering when to move away from your partner to anticipate their hit, like what team UGS does so adeptly. Launchers are the best example, since it becomes so obvious what the optimal followup is. Falcon stomp or rboost, Fox uthrow, Puff utilt or uthrow, Peach dtilt, etc. are all moves that should open a reaction tree for everyone on the screen. As soon as your teammate launches an opponent, run away from them, and prepare for the next hit, such as this: Watch Armada to see perfect anticipation an preemptive spacing. Peach is so slow that, while many of us have the luxury to react to teammate setups and run across the stage to continue team combos, all of Armada’s team UGS ping pong combos are based on preemptive positioning. He knows the distance to put himself at, and is always there for his Sheik to pass to him, regardless of if its Aniolas, M2K, Shroomed, or Android. Making this even more impressive is the fact that he positions himself there while neutral is still being played, zoning out the other teammate at the perfect distance.

Conversely, on defense, as soon as someone launches your teammate, you should be ready to disrupt the followup. This is the reason why so many people are getting saved from rest nowadays, like here: Jab reset or missed tech rest has the same principle: There are many moves that should immediately trigger everyone on the screen to prepare for the next interaction. The next level of yomi comes when you use this knowledge as bait (for instance, I could stomp someone, then instead of going for a knee that would inevitably be interrupted by a Fox running across the stage with a nair or dair, I could pivot fsmash and hit both of them).

Knockback stacking and canceling:

In order to position yourself perfectly, you will need to learn to anticipate when an attack will override the existing knockback and when it will stack with it. If an attack hits someone within the first ten frames of when another attack lands, the knockback overrides. Thus, you can space like this, and the knockback looks normal, nothing interesting happened except damage was added

When a move hits someone more than ten frames after the first hit, the knockbacks stack, often canceling, resulting in strange knockback. Being ready for this can let you get some strange looking combos in the air, and get you some free kills from missed techs as well, assuming you have good anticipation. Thus, if you hit someone near the end of the knockback, it will basically look like it overrides, but if you hit someone before their knockback is late in its trajectory in the opposite direction that they were initially hit, they will likely float there. See and

With electric moves like knee, there is double the hitlag. This means that 6/10 frames required for knockback stacking are occupied by frames before knockback has started. So as early as the fifth frame of knockback, knockback stacking will happen. People getting hit early out of the knee knockback and just floating there is a frequent occurrence, so you should be ready to anticipate this. It sets up for especially juicy punishes, basically making it as if you stomped someone, or as if they got hit off a wall on a stadium transformation with no tech, just helplessly floating there in hitstun.

Grab techniques:

Asides from the basic heuristics when it comes to grabs, i.e. don’t grab approach with grab in 2v2 (landing a grab literally means you lost the neutral unless you are playing a character with exceptional throw animations) and don’t grab in 1v2 unless you want to get killed, there are a number of tricks that can help you really get mileage out of grab punishes. The first thing is to always take kills when they are there. Don’t overcomplicate things because you are worried about hitting your teammate. Just hit both of them and ensure the kill if your teammate isn’t at kill percent. I.e. do this:, not this:

Startups of any throw provide 8 frames of invincibility. So once you have ironed out grab situations with your teammate, you can practice using these invincibility frames to avoid taking any damage. Note that initiating the throw will also change the hurtbox of the opponent who is grabbed.

As listed before, grabs are a situation where you can always eye everyone on the screen. Fox uthrow or Falcon dthrow → attack the other teammate is frequently a good idea. Unexpected target switches get you free kills or crucial saves, e.g. Nothing is worse than a teammate who tunnel visions as soon as they land a grab. Remember, use time as efficiently as possible. Pummel → buffer dthrow is guaranteed at any percent, so you can always squeeze out a little extra grab time if your teammate is out of position by doing this.

Grab releases are another way to guarantee kills. The basic, unforced grab release is a good way to guarantee the strongest hit a team can get given time to wait for the grab release, such as here: Be careful about doing this against floaties, as their aerial grab release can send them significant distances. It’s important to remember to mash most of the time, but sometimes it’s best not to mash, because mashing would give them the time to do a grab release punish when they wouldn’t otherwise have the time. There’s also forced grab releases. Forced grab releases are likely the most ridiculous thing in competitive Melee. Anytime you hit someone who’s grabbed someone else, it results in a grab release, which sends characters out at an obnoxiously low angle. Grab releasing Fox, Falco, Falcon, and Marth often results in kills outright. Do this when the opponent’s back is at the ledge by hitting your teammate, ideally with moves like Falco laser, Sheik needle, Fox dair, etc. Look at how ridiculous this is: That’s a gimp at 16%.

Sometimes, if you know your teammate is going to land a hit but it isn’t going to kill, you can use a grab to snag them out of hitlag, then get another round of hits in immediately. This is especially common in situations where someone up+bs onto the stage and is in lag. It is a good way to tag on a little extra percent to try to close out a kill. Here is an optimal use: Don’t forget that you have almost no lag once whoever is grabbed gets hit out of it, and you can immediately act out of it to punish!

Learn how your hitboxes work with grabs. Spacie fsmashes, for instance, never need to be spaced. If your teammate grabs the opponent and you fsmash your opponent, holding shield will buffer a shield, automatically powershielding the fsmash because the grab will be broken once the fsmash lands. See here: Other grabs can lead to weird phantoms due to hurtbox manipulation. Learn the little cases with your team composition.

Advanced ledge guarding tactics:

It is often said that Melee is a game that as played at the ledge. This is even more true in doubles than in singles. The reason for this is that the natural ebb and flow of doubles leads all parties being at the ledge, whereas in singles, one person often gets cornered, and then that person bides his time, carefully picking his spot to get out of the corner. In doubles, the flow is such that, once one person gets hit off the stage, both opponents collapse on the ledge setting up a team ledge guard, and the other teammate comes to try to disrupt the team ledge guard. A huge meta develops around this, with one side trying to secure the ledge guard while the other tries to maximize the possibility of recovery or return kills.

Ways to help guarantee kills:

  1. Zoning out the support—this is the most important first step. Ledge guard like PewFat does here: (except for SFAT’s atrocious usmash). The worst thing you can do in doubles is to ledgehog as soon as you hit someone with a slow recovery off the stage. USE YOUR TIME WISELY. Zoning out the support ENSURES the ledge guard. Ledgehogging immediately almost ensures recovery of the opponent because you are unlikley to be able to disrupt the save. Ledgehogging as soon as you hit a Sheik off stage is one of the most sure signs that you are terrible at doubles.
  2. Ledgehog and roll if you’re above 100%. In singles, getup attack is often better than roll at under 100%. At both under and over 100%, roll is better in doubles. Rolling from the ledge at over 100% occupies it for a ridiculous duration. You can cover things like Sheik up+b into the ledge and up+b straight up with one roll.
  3. Use additional time to rotate positions with your teammate. After hitting someone off stage or hitting the support away, you may have time to swap who’s on the ledge with your teammate. Don’t settle for resetting to another ledge guard situation. Make sure whoever is going to throw out the attack has the strongest possible finisher. Against characters that struggle to get reprisal kills, such as Samus, Marth, and Falcon, it may be appropriate to rotate the zoning character on stage and have the other character go offstage to clean up the ledgeguard. Going offstage is generally unsafe, but if your teammate can cover you, it can be made safe. Usually people do not properly assess the situation and have poor discretion when going offstage.
  4. Babysit just like in singles. If you are unfamiliar with the term, Mango coined it to describe situations where you are moving to cover an option, but actually still have another option in mind to cover, i.e. babysitting a second option. For instance, jumping high offstage to cover a high recovery, while still babysitting the ledge. Jumping off stage in a way that threatens recovery then double jumping back to ledgehog can be very effective for reducing recovery options.
  5. Push the recovering player’s teammate onto the ledge by hitting his shield. Forcing people to ledgehog their teammate is an extremely effective and common situation.
  6. Use an alleyoop as a substitute for an outright kill. For instance, there’s often a situation where Fox can up+b straight or diagonally up. In this case, instead of doing a difficult reaction with a knee, I can always opt to cover diagonally up non-reactively while letting my Marth ledgehop uair to cover straight, which will always pop him up for the knee. Even in situations where there is only one option to cover, using an additional move can make the timing more lenient*.
  7. Reduce ledge guards to 1v1 when doing a 2v1 ledge guard is not possible. Grabbing the recovering player’s teammate is a great way to do this. Sometimes you have to actively engage them just to create space for your teammate to ledge guard. This is particularly common against Falco, who can disrupt with lasers.
  8. When a Falco is lasering you and you are trying to ledge guard his teammate, teeter or run off jump and tank the laser. The laser will turn you around and you will ledgehog. This makes it so that it takes them a minimum of two lasers to save their teammate instead of just one.
  9. Hit your teammate toward the recovering opponent. It is immensely frustrating when you and someone else get hit off stage, you are slightly further away, and then you eat a bair. There’s nothing you can do. If your teammate is a Sheik, and is around 70% and the opponent is a similar weight character around 90%, hitting both of them off is almost surely going to result in the opponent eating a bair. See SFAT and PewPewU score a cheesy kill here: Here’s some similar gifs to drive the point home:

*The alleyoop philosophy can be extended to other situations as well. In tight timing windows, it is often better to let your teammate do a move, e.g. spacie bair, and then to do your finisher, e.g. knee, right outside of it, because this increases the leniency. Space right outside their range and time your move at the end of hitlag, like I do here In cases where your laggy finisher might be a couple frames slow, you can still guarantee a kill by using your teammate’s faster startup move to combo into your finisher, such as here:

While you are trying to secure kills, your opponents will be trying to save each other, so you will need to learn how to counter many different tactics used for saves. There are three types of saves. There are saves that leave both the opponents vulnerable, ones that leave one teammate vulnerable and the other safe, and ones that leave neither teammate vulnerable. Let’s address these in reverse order.

An example of the last save would be a Fox jumping past his teammate, shining him on stage (giving him no lag), and then still having access to all Fox recovery options as well as his jump. Obviously, this is very bad to let happen. It is also very hard to punish after the fact. This is an example of when you need to be on the hunt. To be on the hunt means to really be after a kill. You should go off the stage and kill the recovering party as long as it doesn’t expose you to a shine spike. If this is not possible, you should try to hit the Fox before he can help his teammate recover. As soon as that shine lands, your advantage is essentially gone. Read the save and punish preemptively. If you can read his movement and open him up before he goes off stage, you can opt for a 2v1 instead of ledge guarding his teammate. Alternatively, if you manage to intercept the Fox right as he’s about to save his teammate, you can turn it into a double ledge guard situation. If you are on point with your read, you can usually get a double kill, intercepting the save by hitting both of them just before it happens. When playing against people who are exceptional at setting up saves like this, always anticipate, do not react, to the situations they create.

In situations where one party is left vulnerable, it may be ok to react. In that case, you are still likely to get a free kill. Saves like this include Puff or Sheik uair, Falco shine, Fox or Sheik up+b, etc. If someone is launched up by their teammate, you can react and still pick off the recovering party. More skilled floatie players will opt for moves like Peach and Puff dair, which do not leave their teammate as vulnerable. In this case, there is a greater urgency to read and disrupt the save, though you still might be able to react and clean something up. Going off stage and hitting a Peach out of her dair will likely result in a double kill, but be don’t be reckless. Going off stage without a well thought out plan and very intentional actions designed to counter specific options is a fool’s errand in doubles.

The last type of save is one where both parties are left vulnerable. There is the least urgency to anticipate and prevent these. However, in most situations where these arise, it’s hard to clean up both kills, but very easy to clean up one. This is because it ends up being a double recovery with awkward timings and recovery angles. It’s not easy to double team ledge guard in these janky situations. See these gifs: With recoveries like these, if you want a double kill, you should read them and disrupt their help. But, as I said, going offstage without a plan is a terrible idea. So it is a bit more outlandish to read these crazy recoveries. Still, in any situation where both opponents have no hope of recovering alone, their only hope is to recover together, so hitting either one of them will prevent both from recovering. It is often ok to settle for one reactive kill, unless circumstances demand you go for more. Guaranteeing one kill like Hbox does here and seeing if you can get more is completely fine

Saving your teammate and regaining stage position:

Knowing when and how to save your teammate requires knowledge and discretion. More often than not, you should drop your punishes when your teammate needs help recovering. Once your teammate is hit offstage, you need to assess how to best help your teammate. Sometimes, positioning yourself near the ledge is enough to save them. A threatening presence can deter ledge guarding. Other times, you should read an unsafe ledge guard (e.g. Marth fsmash) and immediately punish. When someone is between you and your recovering teammate, do your best to engage them to prevent them from ledge guarding. The window of time to do this can be extremely small, particularly against Puff and Peach, who make opportunities to save your teammate extremely fleeting. If you don’t disrupt their combo quickly, your teammate will be carried off stage and comboed into space where you cannot save them. Act with extra urgency if you see your teammate potentially getting comboed off stage by one of these characters. Be extremely careful not to hit your teammate onto the ledge. Getting your teammate ledge hogged sucks, and it happens all the time Grabs are good for holding someone in place while your teammate recovers.

Sometimes, your teammate can make it above the ledge but not on stage. This is particularly relevant with Falcon and Ganon, whose hurtboxes always go above the stage when attempting to sweetspot. In these cases, running across the stage into a teeter gives you many options. When you run to the ledge into a teeter, you can immediately jab, dtilt, etc. Alternatively, sometimes you can just dash attack the ledge. Don’t pick your actions based on the fact that you feel like you have to help your teammate. Pick your actions based on what is actually going to help them. Sometimes it’s best for a Fox to just shoot lasers and get percent instead of fruitlessly trying to save his partner and exposing himself. Even small overcommitments can result in needlessly losing your stock, like this commitment:

If it is not possible to save from on stage, you may want to go off stage. The lower your percent, the less likely it is that you will want to do this. The higher your partner’s percent, the less likely it is that you will want to do this. If you are significantly behind, you will want to do this. Several factors should influence your move selection when going with offstage saves. You want to use moves that don’t leave your partner vulnerable, like Peach’s dair. Moves like Puff’s uair often result in immediate pickups by the opposing team, frequently getting your teammate killed right after getting saved. You should pick the move that exposes you the least after the save—ideally you don’t want to have to up+b after a save, though in some circumstances you will have to. Long lasting hitboxes, like Peach’s dair, are ideal for saving because you don’t need to be as precise. It is often a good idea to pick a move that will horizontally launch your teammate to the opposite side of the stage, effectively improving your team positioning by getting one of you out the corner.

Once your teammate is hit off stage, you should attempt to control as much space as you can adjacent to the ledge. The more stage you can control, the better position your team will be in once he recovers. Sometimes “as much space as you can control” literally means retreating to the ledge and controlling it, preventing ledgehogs. Other times, you can stand your ground and control half the stage, meaning you will not be at any positional disadvantage when your teammate recovers.


Many players lack the discipline to correctly play around invincibility. Committing to a ledge guard or offense in general when someone comes down with invincibility is dangerous. Many times, someone will ledgehog a character like Marth or Sheik, then have no chance to finish out the ledgeguard because an invincible opponent is there, completely precluding the possibility of throwing out the desired finisher. The situation has now completely flipped because you’re in the corner and your teammate can’t do anything to stop an invincible opponent. Though it’s a perfect time to go for a 2v1 when you kill someone’s teammate, the window of opportunity is extremely small unless it’s a star KO or an FD spawn across the stage. Try to be proactive and do something like steal the ledge immediately, and if it doesn’t go your way within one interaction, back off. Here’s an example of good discipline, maintaining strong positioning even when an invincible Fox comes down:

When you come down with invincibility, immediately look to see if you need to help your teammate and if you can punish any lag on moves that the other team may have thrown out. If this is not the case, run towards whichever character is easier to pick on, e.g. run towards the Marth at the edge and try to force him on stage for a double team rather than going to the other side of the stage with the Puff pounding off stage. If you and your teammate have double invincibility, sometimes it can be good to come down together and try to spam kill moves, but other times it’s good to let one person go first and try to herd the opponents into bad position, then to have the character with the stronger finishing moves come down and try to capitalize on positional advantage.

Regaining footing after losing the neutral:

In situations where both you and your teammate are pushed to the ledge, you and your teammate should have a plan for how to best regain stage position. When I team with Fox and he is recovering, I drop from the ledge, let him grab the ledge and ledgedash, then immediately regrab the ledge and get back on myself. If you are not teaming with a Fox, the plan might be more complicated. Be sure to figure out which options are strongest for your team regaining position. Ledgehop aerial and then shielding can be good, as if you whiff, your teammate can attack through you from the ledge. Here is an example of effectively regaining position from the ledge taking as little risk as possible

Shielding is better in doubles than in singles because grab is the counter to shield and grabbing is not good when your teammate is there to break up the grab conversion. Shielding can be very safe if your partner is good at working around your shield. Hitting through your ally’s shield, extending the hitbox, is an extremely powerful to restart offense. Then, he can roll or WD OOS. When in an extremely defensive position, such as on the ledge, let your teammate swing because even if it does not hit, it will force the opponent to shield or move away. Be sure to use light shield and shield angling in doubles, as there is less urgency to get out of your shield, a consequence of which is you are more likely to get shield stabbed. Having a teammate who doesn’t know to hold his shield is one of the most infuriating things because it completely cripples your ability to start offense. If you are acting offensively out of shield, your teammate can’t attack through you with confidence.

Read the pressure of the opposing team by studying their habits. Some people throw out a lot of moves that can be punished, e.g. Fox usmash, so if you are in good position, you can punish them for hitting your teammate or your teammate’s shield. Other teams will crossup shields to start sandwich pressure, as I suggested doing before, so if you can read someone is going to crossup your teammate’s shield, you might be able to outspace them with a move of your own. Occasionally you might even be able to bait someone into grabbing by holding shield, then have your teammate punish them for their bad habit.

Disrupt team combos by picking a time and place to break it up. Most of the time, this means that you should let your teammate get hit once more, then commit to disrupting the second hit. If you try to disrupt something that is already happening, you will be too late. Don’t be afraid to hit your teammate to get them out of a bad spot. Hitting your teammate can get them out of bad situations completely—a Puff won’t care if they get hit by a horizontal knockback move at 70, but they will care if they get hit by a Fox uair. Meteor smashes are great for disruption because they hit your teammate down into the ground, allowing them to tech. Be decisive—don’t flounder like MacD does in this legendary gif: Long lasting hitboxes can be great for preventing kills, like Ice’s drill that saves Armada here:

Invisible Ceiling Glitch:

ICG can lead to retarded kills. If you have a strong feeling that someone is going to hit your teammate’s shield near the edge, read it, and go out of your way to try to hit them off the stage. You can kill people like this, laugh at them, and then watch their mental game crumble

Players to study and other resources:

SFAT for taking space and reading “good options” and countering them. PewPewU for never putting himself in a bad position and maximizing advantageous positions, especially via crossups. Armada for preemptively positioning himself for team punishes while still zoning or fighting the other teammate. Android for combo disruptions. Alan for team saves. L for the most outlandish fucking Fox saves. Other people who aren’t bad to study: Plup for amazing positioning, both defensively and offensively, but especially defensively. Shroomed for offensive positioning/punishes and advanced ledge guarding tactic #9. Me and Reno for dedicated support roles. Obviously there are many more amazing doubles players than people on this list. However, I feel these are the players whose strengths are easiest to see, learn from, and replicate.

Fun teams to watch: PewFat for [NorCal] fundamentals. Armada+Android for optimal punishes and situational tricks. Plup+Hungrybox for Plup’s great teamplay and Hbox’s amazing explosive, aggressive style. Javi+Twin for the most flushed out, balls to the wall double offense there is. Shroomed+S2J for drunken master double offense despite having a terrible composition for it.

Teams 101:

Teams 102:

Smash Lab, Bros before Pros:

PewFat vs Plupbox analysis:

About the author:

NMW is an aspiring Falcon player from NorCal, currently ranked 9th on the NorCal power ranking. Doubles accomplishments include 5th at INY and 13th at Genesis with L, and winning SSS, Mayhem, and Emerald City III with SFAT.

Here are some doubles gameplay highlights:

And some commentary as well:

Contact NMW on Twitter @NMWhittier or @NMWFalcon, and subscribe on YouTube for content.

Teams 102 by NMW: Fundamental Concepts

The average skill in Melee doubles is painfully low. One of the most common misconceptions is that you need a static teammate in order to have good synergy. If you have the knowledge and decent situational awareness, you should be able to team with anyone else who does. I hope that through my guides, commentary, and even gameplay, I can help motivated individuals hone their doubles play so that we will all have better tournament experiences. Teams 101 is a reasonably comprehensive doubles guide, so remind yourself of its contents at This guide will have overlap but is focused on lucidly explaining concepts through prose. See my advanced guide to look for situational tricks to optimize your conversions. I will use strong language in this guide, making rules sound rigid and universal, but just like in good writing, you start by obeying the rules then break them once you are approaching mastery of your craft.

Time is the most important resource. In singles, when you land a hit, you earn position and an opportunity to follow up. In doubles, when you land a hit, you earn position, an opportunity to follow up, and time to allow you to eye the whole screen. Looking at the entire screen is guaranteed but following up is not necessarily. Look at everyone on the screen every time you don’t have a guaranteed follow up and whenever you can manage to do it even when you do have a guaranteed follow up. Grabs, tech chases, or vertical launchers should invariably allow you to look at the whole screen. Many times, after landing a hit, you should immediately run across the stage and 2v1 or help your teammate recover. It’s much better to get a guaranteed kill, like a grab combo, than it is to try to continue a tech chase or other combo with a reset. Helping recoveries should also generally be prioritized over getting kills.

Using time efficiently means you often have to drop what you are doing in favor of another opportunity. Feel the urgency of doubles. Don’t play neutral unnecessarily. Once one person is offstage, everyone should collapse on that ledge because that is the point of contention where important action is about to take place. If you are waiting for someone to land after you launched them vertically, you are fucking up. Don’t wait. Waiting is bad. Unless you’ve hit two floaties high in the air or far off stage, there is always something useful that you can do. Baiting is also much worse in doubles; brute force is favored.

Don’t switch from defense to offense. There’s a spectrum of defense, neutral, and offense. If you are on defense, you are in an intrinsically bad position to start offense. When you are shielding, on knockdown, or on the ledge, you have infinitely fewer options than someone with freedom of movement. If you and your teammate are in good formation and your front line is forced to play defense, this is time for him to rotate out and let the back line start offense. Why would you try to start offense when you are in a worse situation than your teammate to start offense? After recovering from defense to neutral, you can choose to immediately rotate back in. If the back line’s offense managed to be successful, he may choose to push the advantage further before rotating out. Use discretion in these situations—it’s better to rotate out and get your front line back in the action once there is a reset than stubbornly trying to push after the window of opportunity has closed. Notable exceptions to these rules include Fox’s invincible ledgedash and shine OOS when your teammate’s options are too slow. Situations involving shield drops are the only especially tricky times with regards to knowing who should start offense.

Cover options with your teammate. Don’t overlap. This is most relevant in ledge play, where one player should always cover the ledge and the other should always cover on stage recoveries. Start extremely simple to make coordination easy, then expand upon these things by using more complex options when sensible, in the same vein as expansion on musical themes in a well written composition. Add more tricks to your repertoire gradually. Discussing roles and situations ahead of time greatly expedites improvement in this area. Obeying other rules like not switching from defense to offense makes it easier to cover options together because it makes it easier to be one mind. At high levels of play, most of the time someone hits their teammate, it is the fault of the person who got hit for not correctly covering options together. Crossing streams/jumping over your teammate, going off stage for ledge guards, and trying to follow up after your teammate lands a launcher are common things that mess up coordination.

If you are playing a support role, you should spend 90% of your time watching your teammate. When you are in formation, this will let you react to any development. If you taking the lead, you will need to watch your opponents the vast majority of the time because you are playing on reaction to them, whereas the support is playing on reaction to you. Follow previous guidelines about finding time to eye the whole screen. Opportunities for big conversions are frequently dropped because the carry player is not cognizant of what their support has set up, so try your best to minimize this. Conversely, if you are playing support and you are not cognizant of situations that your carry has set up, you are not doing your job. Support players who don’t cover options with their teammate are literally useless.

Don’t break formation for no reason. Maintaining position is of utmost importance, so breaking position needs to be warranted. One person attacking at a time is generally preferable. When only one person is engaging the other team, this leaves the most options open for your team. You can be ready to double team, zone out the other player, save your teammate, etc. A good support should be ready to punish laggy moves that hit his teammate or his teammate’s shield. If you both engage, there are three possible outcomes. Worst case scenario, you both lose your engagements, which is obviously bad. Best case scenario is that you both win your exchanges, but in this scenario, you might be too far apart to team ledge guard, which means neither kill is necessarily guaranteed, and thus it might not even be better than just winning one exchange. Lastly, one of you could win your exchange while the other loses, which obligates one teammate to help the other. This is much worse than your teammate winning his exchange with you there ready to double team and is the same as your teammate losing the exchange and you helping. Don’t tunnel vision on a combo or ledge guard if you are going to expose yourself by doing so. The worst part about doing this isn’t potentially forfeiting your own stock, but the fact that you obligate your teammate to come help you. Regardless of whether the other teammate gets a reprisal kill on you, your teammate will have to try to help you while whoever you killed comes down with invincibility. This almost always results in both of you cornered. If someone comes down with invincibility and both opponents are in the corner with one of them obligated to help the other recover, this can easily turn this into a double kill. If you go off stage for a ledge guard and have to up+b, you are putting your teammate into a Catch-22 situation where he either has to let you die or choose to put himself at risk, exposing your team to a possibility to losing two stocks. Only do this if you are very cognizant of everyone’s position and have the confidence to know you won’t get counter-hunted, or if you want to trade stocks.

When you are against a fast faller, a Marth, or anyone teaming with a floatie, you will have a preference to kill these characters. When someone is at high percent, you will have a desire to kill them. If you are focused on killing one person, you are on the hunt. You don’t want your opponent to sniff out that you are on the hunt. If your prey knows that you’ve mentally committed to killing him, he can make the cost of hunting him too high. For this reason, you should almost never hunt floaties. If they are just trying to stay safe, they are extremely unlikely to be contributing to the action in a meaningful way. It is much easier to try to keep them away and punish them when they try to come in than it is to attack them on defense. If they are out of the fray, they are useless, so there is no need to kill them. Remember, people open themselves up to being hit when they approach, so just punish the floatie once he tries to get into the fray.

Because time is your most valuable resource, it is not worth hunting someone who is putting himself in a defensive position that is difficult to contest. Mask your hunt by moving deceptively and by switching targets whenever opportune. If you read that someone is on the hunt, don’t play reactively. Read their aggression right away and punish them. Top players have the awareness and spacing to avoid getting counter-hunted, but people on the hunt are in poor mental state more often than not. The compulsion to kill people at high percent instead of team comboing their teammate is strong. If you fully anticipate your opponent throwing out a move, you get amazing opportunities to punish. Always assess the mental state of your opponents if someone on your team is living forever. If someone is at 200% and 4 stocks while their teammate is at 0% and 2 stocks, it’s still almost always better to double team the person at 0. Seeing one person with 4 stocks while their teammate is getting to their last tends to be a good situation because it’s probably indicative that you have been going for the opportune target.

Pick stages that play to your strengths or that neuter the other team’s strengths. Poor understanding of stage selection is a chronic problem, at least in commentary. Don’t pick stages because of advantages in singles that are irrelevant in your match. Here’s a little rundown of every stage:

FD is the Battlefield of doubles. It is by far the most common starter. It favors good teamwork because team combos cannot be escaped via platforms, team ledge guards are simplified because there are no platforms (though Fox can do M2K angles), and reestablishing team formation can be near impossible because there is no way to circumvent zoning characters.

Dreamland is probably the most interesting stage to talk about. It has a lot of ground space, a high top platform, plenty of room to move around on platforms, and huge blast zones. The boons for mobile characters often outweigh the survivability buff for floaties. Living longer is obviously good. However, if you are recovering more than you are on the stage, this is all time that can be used to double team your teammate. Sometimes it is better to not kill a floatie and to take advantage of their slow recoveries, intentionally not hitting them with finishers so you can get additional time to double team their teammate. Having an incredible amount of space to maneuver also makes it easier for fast characters to pick their spots against slow characters, and most importantly, it lets fast characters reset to neutral with ease. If you are playing a fast character and want to stay out of the fray while your teammate plays neutral, this is ideal because you can avoid engagements so easily. With all that said, it can still be good for floaties because of blast zones, particularly if your team composition is not susceptible to someone being double teamed during lengthy recoveries. Wispy complicates team ledge guards because you can get pushed off platforms and have no lag, buffing characters with a lot of up+b lag.

YS is a shitshow but it is by no means random, nor should it be neglected. Despite me being a Falcon player, my teammate and I almost invariably pick this stage. Having to fight in close quarters with your teammate makes it favor teams with better teamwork. Team coverage on platform tech chases are ridiculously easy here because of small platforms at low heights. Randall and wall jumps significantly alter recovery possibilities. Small blast zones buff characters with strong kill power and nerf those who rely on surviving as a strength. Hitboxes cover way more of this stage than any other, so being able to hold and control space is hugely important, thereby buffing characters with large hitboxes. It does offer a top platform, which can be used to find breathing room and to disengage against floaties. It is easier to close distance here, so if you are getting camped and want to force them to fight, consider Yoshi’s. Double Peach and Peach Sheik are teams you want to ban this stage against. Marth’s sword can control a lot of this stage, but it can be really challenging for him to not hit his teammate.

PS does not loan itself to strategic advantages as much as other stages. When double Fox picks Stadium, it’s a buff to their muscles and does not indicate any strategy, meaning they are unlikely to have a plan like “smother them” or “camp them.” Stadium gives transformations, which favor Fox, and which also favor big swings. If you are at high percent, it can be a very good idea to push and try to establish an even stronger advantage. You will have little to lose and a lot to gain because walls destroy the correlation between losing a stock and having high percent.

Battlefield is perhaps the least interesting stage. It has a high top platform, allowing characters with good vertical mobility to choose their spots carefully. It nerfs Fox recovery. It is the only middle sized stage, so if big or small doesn’t sound good, you should go here. There is only one small stage in doubles, so if you want small and can’t pick Yoshi’s, you should go here.

People are painfully bad at 1v2. To be good at 1v2, you need to be as threatening as you can while as noncommittal as possible. Don’t over commit on the person at high percent and don’t over commit on ledge guards that expose you once you finish them. Move a lot. Moves have lag, movement doesn’t. Try as hard as you can to be slippery and to make it unclear which opponent you will attack. Being predictable will get you locked down. Once you aren’t moving, you are probably dead. Crossups are especially good in 1v2, though you should be wary of potentially putting yourself in sandwich situations. It should go without saying that you want to avoid shielding. And for the love of god, don’t grab. A good way to practice 1v2 and 2v1 is to play with three players. Have one of them on a team with one stock each while the other player starts with three stocks. This definitely favors the team. It forces good teamwork, and once the team has decent teamwork, it will force you to be slippery in order to beat them. Here is an example of a good 1v2: Note that Marth has fast throws with favorable knockback angles, so he can sometimes get away with grabbing in 1v2.

About the author: NMW is an aspiring Falcon player from NorCal, currently ranked 9th on the NorCal power ranking. Doubles accomplishments include 5th at INY and 13th at Genesis with L, and winning SSS, Mayhem, and Emerald City III with SFAT.

Here are some doubles gameplay highlights:

And some commentary as well:

Contact NMW on Twitter @NMWhittier or @NMWFalcon, and subscribe on YouTube for content.