• Welcome to Smashboards, the world's largest Super Smash Brothers community! Over 250,000 Smash Bros. fans from around the world have come to discuss these great games in over 19 million posts!

    You are currently viewing our boards as a visitor. Click here to sign up right now and start on your path in the Smash community!

Comprehensive Amiibo Training Table

Comprehensive Amiibo Training Table

Applicable Games
Smash 3DS, Smash Wii U
Amiibo Training Table

•I have an in-depth breakdown based on guides already out there, but is a more intricate level-by-level training.•

[--Truthfully, the purpose of leveling in smash bros. amiibo is equipment. Can't feed your amiibo more equipment if it doesn't level, can't feed it when it quits leveling. That's it.
--the other purpose of improving based on level is to loosely knit all of its knowledge together into one comprehensible play-style. That's what my in-depth breakdown unravels.
--amiibo level more by losing, so in the case that you put equipment on your own, it is best to let that sucker win so that it levels less, ideally one or two levels only and, consequently, is able to receive more equipment.
--i'm interested in any feedback, ideas, strategies, and techniques that your amiibo acquire based on your own formulas for creation; *I'm also looking for shield damage data, and information about mashing out of grabs in the new Smash* so if anybody wants to make a post for those subject matter, that's on the table. *Raises eyebrows suggestively*.

There are four stages that follow the introduction.
Stage 1 is Mechanics/Stage Position
Stage 2 is Boxing/Juggling
Stage 3 is Zoning/Habit-Forming
Stage 4 is Reads/Habit-Removal
Stage 5 is AT/Refining
•Stage 6 is the post session


–Amiibo Philosophy–
I found some very interesting information in a post elsewhere on smashboards. Deserves to be here:
Genetic Algorithms are, effectively, a program that runs once, manipulates a few values, then runs again with the new values. Basically, things are given "weight", which is used to mathematically determine the course of action. If the A.I. runs and something doesn't work, it's weight goes down. If it does work, it's weight goes up.

Here is an example genetic algorithm. Cars are randomly generated and run a track. Future generations will be more like the cars that get further, meaning the code "evolves", just as life does, by passing on its "genes", or, in this case, weight.

An Amiibo doesn't need to store it's moveset, or any calculations, it just needs to store it's genes. The A.I. is already on the disc, but the values passed to it are given to the Amiibo.

Now, for a bit of math. A floating point number is 4 bytes long. A single precision floating point number can store a number between 1.175494351 * 10^-38 and 3.40282347 * 10^+38. This isn't an even distribution, as you start to lose some precision for very small and very high numbers, but what it comes down to is, a single gene can have 18,437,736,874,454,810,623 different values. Our Amiibo can hold 64 of these floating point numbers, meaning that the total number of distinct behaviors an Amiibo can have is 1.1800152 * 10^21. That's 1,180,015,200,000,000,000,000. That's One point one-eight sextillion different Amiibos. That's roughly equivalent to the number of stars in the observable universe.

You can do a lot with 256 Bytes.
I've noticed that you can influence your amiibo's behavior, in much the same way it will influence yours. This is best explained with a projectile, but the principle carries over to something like a rapid jab lock. Ok, when your amiibo begins to run to the other side of the stage to use its projectile move, such as diddy with his peanut popgun, you should begin to shut it down with the same move, until he learns to capitalize on this maneuver in a more precise way. The amiibo will begin to try something else in its arsenal to complete the task--the most fundamental task is to survive-- and as for a rapid jab, as you begin to outmaneuver the amiibo spamming this move, it too will learn to do the same, and should you capitalize on that mechanic, it too will learn to do the same should you fall into the same situation. They will even use the attack that has the earliest hitbox and/or use smashes that deal the most damage (when they can). This is why they learn to do a smash Attack when you fast fall but never seem to do enough air attacks. Hopefully, you begin understand the underlying philosophy behind how that works, as you train. Another example is that you DO INDEED have to use the mechanic in order for them to learn it, as in their horizontal recovery, ledge attacks and the like, should you pressure them in another way and never let them use the ledge Attack, they never will. But like a guide or two has stated, they do NOT copy you, but instead you "carve" or "graft" their playstyle.

–About habits–
One personal technique I use from level 1-20(this is debatable) is using only food and fairy bottles set to low. Perhaps a variant is to also include containers such as capsules and possibly even explosive containers. This is enough time for them to pick up your habits with these items. Amazingly, they will learn to throw the fairy off the stage if the opponent is over 100% and Attack with it below 100%. It is even possible to train the amiibo to item hog the fairy bottle until it is strategical to pick it up.

Amiibo take into account many different things when acquiring habits like game settings, game type, match layout, and even how far you are along in a match. I'll be interested to see any unique combinations and habits that come up.

Amiibo tend to reflect what happens to them when put into different situations. This may change depending on whether they fall towards/away from you. This may prove to be a good basis for perfect shielding when falling into an approach, and for spacing should you choose to fall backwards.

I've heard tales of people putting together unique combinations of match settings prior to training their amiibo to get them to be defensive, offensive, or rounded. If anybody has any they'd like to share, here is a good collective for these strategies. These are most-likely useful in beating out specific habits and don't affect overall playstyle to a great degree. One example may be to put a handicap of a specific damage to habitualize a kill throw. >confirm?

A very strange thing is that a fairly well-taught amiibo who have not yet learned all of their moves and options will do weird things that are counter-intuitive and unproductive. This is when drilling advanced techniques comes in handy. If your amiibo goes to the other side of the map and begins to jump, it's a sign that you should 1) run over to him, out-maneuver him, and one-up him in a long, technical, professional level trade that you've done a million times or 2) you can do this a very easy way and go straight to the (either) ledge. Watch how they prance around dumbly, and if they don't have an attack that covers edges well they will continue to not understand the mechanism. There are many options on-stage if you have a projectile, but those require precise aim, skill and timing--including playing a game of catch**.

**you and your high-level amiibo can play a game of catch to infinity, or at least, your amiibo can. Were you to force the amiibo to catch an item, throw it at you, and it hits you, the next time this happens, the amiibo will always choose to throw the item. Removal of this habit leads to bigger and better outcomes for your amiibo as this prevents the amiibo from improving. This is because the amiibo tries to outdo you by damaging you the same way it worked the first time (they get nasty with their maneuvers), and if you take damage it sees that as a good sign to continue that habit. On the other hand, if you always evade its throw then you are not outmaneuvering your amiibo but defending, causing your amiibo to create a thorough, yet technically bad habit.

An amiibo can acquire tricky, though sloppy habits based on your own, and are essentially imperfect. To perfect the amiibo you have to understand that instead of choosing to KO you, it is attempting to outmaneuver you based on its own grafted (not copied) playstyle and THEN get the KO (usually caused by a sloppy trade). This is ineffective when training amiibo vs amiibo. This is not always the case as I have seen killthrows pop up without managing to sink many, if any.

A concise, clean move is to allow your amiibo, after reaching the final level, to train against each level CPU so that it picks up their good habits. This way they don't pick up their bad habits halfway through training. Oh, what a buzzkill to have a solid understanding of how your amiibo has been learning and have to backpeddle, spending your time beating out the dumbness of programmed code. Having been taught by this guide, by the end of training your skill level should allow for the amiibo to have a response for everything a lvl 9 knows and much more with minimal faults. Watch for any techniques it hasn't learned such as stage-teching or respawn privileges.

–Removal of Habits–
Amiibo pick up habits as they level--some of them their owner may never be able to remove--this is based on knowledge of your character, skill and how good you are at ditto matches. The most efficient way to get your amiibo to most closely resemble you, so that when it does finally match your skill level, is to learn to outdo what he's been taught and remove habits that are inefficient such as an improper throw.

The most pristine way to remove an amiibo's inconsequential habit is by using the move/maneuver against it. Outdoing Diddy Kong popgun to popgun allows him to learn to approach and/or use his banana. To remove a snuffy Falcon Kick, you have to block it with a Falcon Kick of your own so you amiibo tries something new.

Outdoing and teaching your amiibo THROUGH the basic skill known as the damage game removes imprecise habits that are inferior yet tricky.

The most pristine way to remove an amiibo's CONSEQUENTIAL habits involves patience and proper timing. For instance you've been bullying your amiibo to go to the ledge during training, but before it does this it decides to shield drop from perfect shield attempts to a raw shield. This is dangerous because depending on the opponent, it may pick the raw shield off before the amiibo has a chance to react. This is easily avoided after a decent trade in which it is useless to proceed--stop doing anything and see exactly what it is your amiibo does. This is useful for allowing your amiibo to get creative with its actions, fight its way out of the corner, and possibly opens new doors. Conservativism involves training your amiibo to go to the ledge, while being more liberal allows for the amiibo to metaphorically dip the toe.


Level: 1~5
Stage 1
has three steps. Step 1: Go straight to the ledge using quick ledge grabs, as it is a very long process before the amiibo learns just what to do and where to do it. It tries every possible way and in every manner to remove you from the ledge, but not before fooling you into using one of the ledge options. If it succeeds, it will try to "snuff" your ledge getups over and over with different moves until finally trying something new. To accomplish successfully teaching your amiibo this mechanic, use your ledge options to bop your amiibo when the opportunity arises. This takes patience as you will be hanging from the ledge for quite some time. Make sure any attempts to attack you on the ledge are used especially if they don't work*. Eventually the amiibo will ledge trump you from the stage. This is when it gets tricky, as you need to successfully remove Ill-timed ledge get ups from the amiibo. The same process applies, but in reverse. For example, after successfully using a ledge Attack, it may wait a moment and attempt this again, so here, your job is to allow this until hitting you with a ledge Attack is unreasonable. Then begin to challenge its ledge options, and after, the next step is to stage-spike him until the the amiibo learns to avoid it. Badger it at the edge until it learns how to time his invincibility frames and avoid free damage. Step 2: the amiibo needs to consider shield, both as an option for itself and as a potential game-breaker by the opponent. For that you need to raw shield. Allow the amiibo to break your shield so that it learns to rushdown+shield pressure. This habit will be removed later in training. If possible, attempt to break the amiibo's shield as well, rather than grabbing the amiibo out of it. Step 3: practice wall-jumps, teching, and footstools, so that your amiibo has a developed understanding of these mechanics and can foresee these things where you may not. That should cover that, but don't remove it from your gameplay.
There are five main stage positions:
1) center of the stage
2) the ledges
3) the position on top of a ledge/the edge of the stage
4) the position to cover a roll-getup from the ledge
5) the ledge guarding position on the opposite side of the stage

For Stage 1, 2-stock matches with no time limit is standard as you are going to be practicing a little while.

*don't die while on the ledge or the amiibo will see this as a sign that whatever he is doing is effective.


Level: 5-15

In Stage 2, stay within your amiibo's zone and land (mainly) smashes and ground combos. Roll around them and flesh out this process. Don't let them escape, and jump away. Make sure you are applying shield pressure. Refrain from dash in this portion, stick to walking/skating/rolling. When this no longer does the trick, sink guaranteed combos from grab until they react. (Pull out your preferred item and practice comboing with its properties. For instance my first amiibo is Diddy and it is absolutely necessary to master the banana toss at the start of the match as it is key to sinking a full combo. And for those who use amiibo with tether grabs, Avoid them at all costs). You need to achieve full damage with each low-percent combo.
I know 65 percent is a common full combo, but an appreciable combo can nearly always land the opponent in the mid-40%'s.
Then, juggle them. Give them no stage position. Ledge trump them. Standing still is unnecessary here and can be replaced with safe jumps or dash dances, but these techniques are more important in later stages that involve spacing/timing/tech-reading. As for projectile users, don't be overly aggressive. When you reach your boxing/juggling limitations, spam your guns. This should should be repeated in your preferred manner, until your amiibo reaches level 15. DON'T MISS. Practice on FD only. Win with KO moves/throws/setups. This is crucial for effectiveness.

Level: 15-25
[Port Priority: now is also the time to switch your amiibo to P2 so that in the event it grabs you at the exact frame you do, you are awarded the grab.]
Add a time limit. In Stage 2, stay out of the amiibo's zone. Capitalize on any lag it admits. Force it to approach. Make sure your moves are at least clashing with the amiibo's. Interrupt it's recovery, stage-spike, meteor--anything that makes it try to recover high (but don't overdo it with jab finishers(; ). It is the time to put a spin on your amiibo (via strategies others have posted or your own) if you wish. You need to stay dedicated to your particular strategy until until the amiibo reacts. For instance, if you put a handicap on your amiibo or yourself, you should keep that strategy live until you find yourself struggling at each of the stage positions. When forming a habit, take into account your spacing and how far along the match is. *This is important because amiibo take damage, stock-lead, and time remaining into account when learning individual habits.* Your Samus amiibo needs to have a reason to spam missiles, for example, so make sure to predict when they roll backwards and reverse zone with the missile spam. The principle to this is that when your opponent is at a certain moment within the match, the amiibo makes decisions to use strategical setup/maneuvers that put it ahead by stock, tell it to react to being behind by a significant % in damage, tell it to roll evasively and so on. If Samus has learned that too far along in a match, say kill percent, she may whip out this fat missile spam and get KO'd by one well-placed aerial. This would be the time to teach it to roll evasively for a charge shot. The opponent may just stand there waiting for you to pop out and do something stupid. You're better than that. Even better is to try this setup off of a jump. Amiibo setups ftw
Practice on tournament legal stages--come back to FD later.
During levels 20-35, make sure you have been practicing proper DI and Smash DI.

{Reads/Habit Removal}
Level: 25-40
Switch to Time matches and play for a length of time. Reading and habit removal goes hand-in-hand. For this stage, stand still and tech-chase with techniques like dash dancing and evasive rolling, etc. Stay out of your amiibo's reach using techniques like pivots/pivot grabs and skating. Style on 'em. If there is anything you want the amiibo to forget such as projectile spam, make sure you use the maneuver against it, forcing it to try something new. If I use the banana, Diddy Kong starts to use the popgun and vice-versa. As for the raw shield, now is the time to run up and grab the amiibo. *Note: when removing a habit, you typically have to use the same move that it is using--see Removal of Habits in the Introduction.*
During levels 20-35, make sure you have been practicing proper DI and Smash DI.


Level: 40-50
In Stage 4, tourney rulesets must be used, two-stock matches with appropriate time limit. try to win by "time" if you need to.
You need to practice any advanced maneuvers--quick ledge grabs, b-reversing with success, attacking from the ledge (different from ledge attacking), bait and tackle, pivoting with success, footstools. Force out their rolls/air dodges and practice punishing accordingly. Auto-cancel your attacks, practice missed-tech opportunities your amiibo present, and round out your performance in any given match. Continue habit forming. By this point it is important that they can stage-tech and use any special-move items they carry in a 1) efficient and 2) effective manner. This is the time to practice choking and edge-guarding.

{The Endless River}
Level: 50
As is generally believed, amiibos continue to adapt after level 50. Perfect 1-frame punishes on ledge-get ups, rolls, and the lag on attacks, continue habit forming and put them in a variety of matchups on a variety of stages. Put equipment on the player character and outdo the amiibo unfairly, THAT'LL teach 'em. Disregard stages with ridiculous gimmicks such as Wily Castle's Yellow Devil. Throw in items, make all your amiibo participate in 4-or-m0re-player matches, brutalize a new amiibo with a level 50. I am excited to see what that looks like with my new strategy now that it is developed.

&Gripes are appreciated&
$Particularly tough habits to form make for a great discussion$
<tips, tricks, interesting maneuvers are also welcomed>
First release
Last update
5.00 star(s) 9 ratings

Latest updates

  1. Data hoarde

    http://m.youtube.com/watch?v=oKt5YxqOWR8 OOOOOOOOOOOOOOOOOOOOOOOOOOO I may have discovered...
  2. Revamped and Updated

    I have edited and reiterated my guide to a great degree. I stayed up late, I lost all my work, I...
  3. Update after update

    i've been editing this thing like trimming a base full of bushes, and am still. I am planning on...

Latest reviews

Excellent guide! I've been using the method in this video:

I'm gonna try your method and see how the amiibo compare when facing each other. Thanks for making this!
Very nice. I want my sonic amiibo to be the best
This guide is very interesting because it has much more detail.
Sadly my amiibo is levelling up too fast before I am able to teach it everything in this guide.....
Where's the hyperbolic time chamber when you need it...
Thanks alot. I couldn't find any other decent guides out there. I love how in-depth this is and the effort put into it shows!
Thanks a lot I've really learned a lot since I first wrote this in, and I think the biggest obstacle now is a good teammate amiibo.
Very helpful
super in depth and fun :)
Super.. Thanks! I really hope a bunch of people add to the topic in the comments. It'll be nice if we can see a unified collection of ideas and strategies. If there are any discrepancies as well let me know.
Top Bottom