Thursday, August 11, 2016

Intro to MangoPuff: Profiting From Risks

Intro to MangoPuff: Profiting From Risks
A guest post by Eru

In this article I'll be analyzing the neutral game of Mango's Jigglypuff vs. Fox with the goal of answering an important question: What can Puff players (and everyone else) learn from the neutral game of Mango's first main, specifically in the Fox matchup?

My answer is simple: quite a bit. But first, let me explain my method. I'll be drawing my examples in this piece from sets vs. Jman, Mew2king, and PC Chris*. I've chosen these players because at one time or another it's fair to say that they've all played the Puff matchup well. The more recent material from Mango's stream (vs. Lucky, Alex19, and S2J) is helpful, but not really appropriate for an introduction. I will use these examples to illuminate what we’ll call the two rules of MangoPuff:
  1. Getting hit is a good thing.
  2. Unsafe aerials are a good thing.
Obviously both of these principles need some explanation; so let's get to it!

Principle #1: Getting hit is a good thing.

For the rest of this piece I'm going to talk about the neutral game as a data-gathering mission. Every time you put out a move/jump/etc and elicit a response that response is valuable data. Ideally, once you acquire enough data you will be able to predict the opponent's response to a given action or situation and can maximize your profit on confirmation of their commitment. This model works very nicely for MangoPuff, particularly with the application of principle #1.

Simply put, MangoPuff prefers to hang close to the opponent and trade hits. This is because each time he hits or is hit he gains information. Take the opening of this set as an example. Mango nairs towards Jman three times and fairs at head level three times. Why? This is obviously unsafe and could be cc-grabbed, upsmashed, etc. The answer is simple: data. Mango is, in effect, asking Jman, “What do you do when I approach?" Jman's responses over the next few minutes determine Mango's course of action.

Moreover, it's also important to understand that principle #1 acts as a way of keeping the opponent honest. Think about it; are you worried about an opponent who never hits you? Of course not. When MangoPuff becomes a pest to the opponent, hanging around at head level like a swarm of bees, he not only gains data from that opponent, but he forces the opponent to respect his space.

Perhaps the most important aspect of principle #1 is that by the time you get to the end of a game you have so much data that you really don't have to fear the "random" upsmash because you will recognize when it's coming. In this example, even though Mango is at death percent, he nairs into Jman, fairs his head and grabs him (because Jman shields, a reaction Mango clearly anticipates based on past interactions), and nairs him several more times before ending the stock. There's plenty that fox could do, but only so much that this fox will do.

Near the end of the match, almost everything Mango does to Jman is technically unsafe, but in context it is not. Mango has hit Jman enough times by this point to know exactly what he's willing to do in most of these situations. But more importantly, Mango uses that knowledge to inform his play.  In the second game he immediately hits Jman twice with fade-back aerials. The reason? He has the read that Jman is attempting to dash dance and grab when Puff lands. This knowledge makes it easy to play spacing tricks on his opponent, especially with Jigglypuff’s aerial mobility. But what about getting hit being a good thing? Well, in this example, Mango is hit by Jman’s first full hop dair but reacts to and grabs the second, transforming a "safe" full hop approach by fox into what could be a rest.

Note: if you watch the entirety of the set linked in the examples, you'll see Mango get away with things that your puff/character could never get away with most of the time. The easy (and wrong) conclusion is something about "top player effect, blah, blah, blah." In fact, the reason Mango can do absurd things like this is that he has collected enough data to correctly create, predict and react to that situation. This brings us to principle #2.

Principle #2: Unsafe aerials are a good thing.

Most of the groundwork for explaining this concept has already been laid, so I'll use specific examples to explain it in more detail. When I refer to "unsafe" aerials I usually mean nairs and fairs that could be punished by cc grab, upsmash, a preemptive nair, etc. Mango uses nair and fair as spacing mix ups to allow his puff to achieve the equivalent of a Marth dash dance (Game 1, 10-12 seconds, 38-41 seconds  ) by subtly shifting his perceived threat range. Notice how in each of the cited clips Mango mixes up his spacing by feigning to move forward with nair or fair while anticipating a rush in and perfectly spacing a fade back such that he has space to react and punish. In one case it leads to a kill, in another, a whiffed grab that with better execution could have been a kill. Thus, an "unsafe" aerial is with intention actually clever micro spacing calculated to coax the opponent into a bad situation. Additionally, "unsafe aerials" are a calculation on Mango's part that in many situations it is preferable to trade hits and lose than to leave fox free to take stage, whittle away your options, and finally read an increasingly predictable fadeback.

Diving a bit deeper into principle #2, let’s take a look at the Super Champ Combo set vs PC Chris. This set is instructive for several reasons, perhaps the most important of which is that PC Chris plays this matchup in 2008 remarkably like many of the better foxes in 2016: the full hop aerials, the smart shield/spot dodge use, and the lasers are eerily similar to Armada's approach to the matchup today.  This match is notable for our purposes because PC plays patiently and only takes what he thinks is mostly guaranteed, though Mango will force him to make several plays he would like to take back. In game one, PC almost exclusively shields whenever Mango attempts to poke him with a nair or a fair (Game 1, :29, :48, 1:02, etc), and as a result, Mango gets frustrated with PC's distinctly un-M2K reactions and adapts poorly, losing game one, but not before downloading PC's reactions to several important interactions.

Games 2-5 are a bit different, though in game three Mango is clearly frustrated by PC's unwillingness to fall for his "traps." Hungrybox normally expresses his frustration in-game when he mindlessly bairs in the same pattern, allowing his opponent to wall him off or punish a telegraphed land/fadeback. When frustrated, MangoPuff will make this sort of decision. A badly placed nair that lands directly in front of Fox is easily punished and confirms for PC that he's beginning to wear Mango down. The rest of game three sees PC grow increasingly confident against Mango's approaching nairs and fairs, resulting in aerials out of shield, grabs--even a rare upsmash--while Mango is slow to counter-adapt. Slow, but not incapable. While it may seem as if PC is firmly in control of the matchup, Mango demonstrates in game 5 that PC's newly-discovered aggression will hurt him (Game 5, 0:14-16 and 0:53-55). In these early examples, PC survives but only thanks to a few well-placed spotdodges. Mistakes like this on PC's part eventually lead to the beginning of the end in this exchange (note the response to PC's full hops).

Let's summarize what we've learned from this short time with MangoPuff:

1. Mango prefers to risk trading hits over surrendering space to fox because it allows him to gain information. Information is king. 
2. One of the best ways to get the sort of information he's looking for is through "unsafe" aerials and approaches.

As an addendum, let me add that by recognizing when and where the opponent will try to whiff punish an unsafe aerial we've also been provided with a functional model for Puff's version of a dashdance and where it is most profitable to use her fade back. Finally, I think it's important to respond to the inevitable "he's just doing good player things" charge that is most commonly leveled at Mango's Falcon/Puff/Marth. The proper response for the thoughtful melee player: so what? Seriously, why is watching a good player do great things with your character somehow inapplicable to the development/theorycrafting that happens with that character? Let the "good player things" be a glimpse into the alternate potential of a character rather than an anomaly.  Melee isn't baseball. A small sample size shouldn't be immediately discarded or overlooked; rather, the nature of melee demands careful qualitative analysis that can yield impressive results--PPMD and 2016 Hungrybox are recent examples of the value of this approach.

There's so much about Mango's Puff that's still woefully neglected (run up shield and cross ups spring to mind), but hopefully this initial foray into MangoPuff analysis leads to a fuller understanding of neglected tactics with much to teach us.

* I would love to use material from his stream against Leffen last year, but it seems that no one saved that recording and it is forever lost to time. If, by chance, you happen to have/know where that recording can be located please don't hesitate to share.


  1. Great read! How would you say mango's puff in 2016 (in friendlies from his stream) compared to his puff back when he mained her.

  2. puffgamesHere are the most popular online games to play for free. Puffgames is the best place to play high-class browser games; Also you can find all most played game on it.