Unnecessary Limitations

One of the longest-running issues in the Smash 64 competitive metagame is that there simply aren’t enough tournament-legal stages available for play (while Battlefield and Final Destination did exist in the game, they are only accessible through Gamesharks, which aren’t regularly available anymore). For the longest time, the only stage allowed was Dream Land, which is unarguably the least intrusive stage in the game. But the community had a problem: characters that performed better on Dream Land had the advantage. By not having any other legal stages, character choices that otherwise would have been useful options were left in the dark. The metagame was Dream Land-focused, and it showed.

The community solved this by compromising. Hyrule Castle, which is not conducive to fair play in certain areas, and Congo Jungle, which is really unconducive to fair play against certain characters, were eventually allowed. This balanced out the metagame and allowed for more viability, and frankly, more interesting matches.

Amiibo tournaments follow a similar issue. Most amiibo tournaments pick from two options: Omega stages, or Battlefield. Occasionally they will go with both. It’s understandable that the tournaments would do this because Cloud Nine, the current king of the amiibo scene, recommends that all matches are played on Omega stages, (let’s be honest, and no disrespect to anyone, the amiibo scene is quite a tiny kingdom).

640px-SSB4UPilotwingsOmega.jpg

The problem is Omega stages don’t have much on them. By definition, they’re flat stages. And as with Smash 64, characters that do better on this type of stage are going to do better overall, due to only having advantageous stages to play on. These characters include:

  • Little Mac (he was designed specifically to be excellent on the ground but not in the air)
  • Ness (amiibo would have to jump, shield, roll or dodge to avoid his PK Fire, and if they don’t jump he can dash grab them)
  • Bowser (especially in customs/nonvanilla, with his Dash Slam custom move that covers a lot of ground)
  • Marth/Lucina (Dancing Blade is a horizontal attack, as well as their mighty tipped forward smash)

There’s obviously a few more characters that would perform better on flat stages, but these are the most relevant to both the vanilla and customs/nonvanilla metagame.

So what can be done about this? Well, some tournaments haven’t only gone with Omega stages. CNAL I went with only Smashville, which was helpful but didn’t completely solve the problem. Some tournaments go with Battlefield on the first match, Omega on the second, and Battlefield again on the third. That is also helpful, but it too doesn’t completely solve the problem. Obviously these solutions are well-intentioned, and I can’t fault anyone for coming up with them. But there may be a better way.

Let’s expand the allowable stage list, and choose the stage with the game’s Random Stage feature. We should change it to:

  • Battlefield
  • Final Destination
  • Miiverse (even though it’s the same as Battlefield, Battlefield is such a useful stage that there’s little reason to not give it extra representation)
  • Smashville (Smashville is possibly the most balanced stage in the entire Smash series, as it neither favors platform-based characters nor ground-based characters)
  • Dream Land 64 (this stage is contestable, as it has a legitimate stage hazard that could interrupt play a small amount. This stage should be up to the individual tournament)

By having a much more diverse stage list, we solve the problem of most stages being advantageous for ground-based or platform-based characters. For platform-based characters, we have Battlefield and Miiverse. For ground-based characters, we have Final Destination or another Omega stage that the TO prefers.

Smashville and Dream Land 64 are the in-between stages. Smashville is a ground-based stage until the platform comes back to you, which keeps it from focusing on one style of play more than the other. Dream Land 64 is large enough that it acts as a ground-based stage until someone is knocked vertically, at which point it becomes a platform-based stage. And by having the game itself pick the stage, we can ensure as fair a representation as possible of the stage types.

Changing the stage list also helps buffer against a less noticeable, but more lethal issue. The amiibo metagame is dead. Not many people come together to participate in vanilla tournaments (although there’s certainly a few people still doing it on some closed Facebook groups), and the nonvanilla/customs metagame has been settled for quite a while now. In fact, one could easily make the argument that only way to keep the nonvanilla metagame alive is by continuing to ban or nerf characters or equipment, just to keep things interesting.

By expanding the stage list, we are solving the problem of some amiibo being unable to perform as well. On top of that, we could protect and prolong the precious little interest in amiibo tournaments until Smash Ultimate arrives. Let’s get rid of the unnecessary limitations.

 

Amiibo Science: What could be possible with the Brain Transplant, and what we have done so far

In my last post, I detailed how to take the functional data of an amiibo and place it into a different character. Since then, I have taken some time to experiment with this method an uncovered some unusual results.

Before I dig into this, let me state something very clearly. Brain transplants are not a replacement for high-level training. If you didn’t train your amiibo properly before the transplant, it will still be a bad amiibo after the transplant. In fact, it could be a bad amiibo afterwards regardless of how well it’s trained! This is not a perfect science.

There are limits to brain transplants. Immediately after the transplant, the amiibo will use the same moves that it did as its old character: taking a Sheik (who almost exclusively jabs) and putting him into Mario will cause the Mario to use his jab incessantly. However, over time, the Mario will succumb to his new AI. The Sheik experiences will be diluted and over time will be replaced by the tendencies that a normal Mario would have (down-smashing very often). This is to say that brain transplants are not permanent.

If they’re not permanent, what use do they have?

Well, some of the previous tendencies of the amiibo will still be there. Even fifty matches later, Mario will still throw out jabs when it would be uncharacteristic for him to do so, just not as often as when he was a Sheik. (I’ve tested this, too. I stopped testing at fifty matches because the effects just wouldn’t go away.) Mario will still do what a Mario does and down-smash often, but he won’t only down-smash anymore.

If these tendencies remain in the amiibo after the transplant, then we could use them to make other amiibo better, too. I recently trained a heavy Mii Swordfighter, using a 3221 moveset. I taught him to use his neutral special and his forward smash. Being a Mii Swordfighter, he really wasn’t all that great. Swordfighters do well in vanilla, but they aren’t much compared to a human player.

When the Swordfighter hit level 50, I transplanted him into a Greninja. I didn’t expect much, if anything at all. Greninjas are obsessive about their up-smash and almost nothing else, so they have never been a useful amiibo in any capacity. The character itself lacks KO power, which leads to some particularly boring matches.

So after transplanting, I took the resulting Greninja and had him duke it out against progressively more difficult amiibo (a few of which I had trained, but most of which I had obtained from competitive amiibo trainers over the months that I was part of their little group). The transplanted Greninja is roughly a mid-tier amiibo now. This is compared to most naturally trained Greninjas, who are completely useless against other amiibo. Before this, Greninja was a strong contender for the worst vanilla amiibo possible.

(Granted, the transplanted Greninja still has the 3221 custom moveset for Greninja, but I don’t believe that has affected him much.)

Outside of the potential this has for vanilla training, there is also the possibility that this could rock the customs metagame as well. However, I have not been a part of that metagame for a while and am largely unfamiliar with the direction it has been going. Thus, I will leave it up to the readers to test out how this could play out.

Amiibo Science: How to Give Your Amiibo a Brain Transplant

You will need:

  • A physical amiibo figure OR an amiibo Powersaves with Powertag OR an NTAG 215 chip. Regardless of which one you use, it should already have been formatted for use in Super Smash Bros. The character on this hardware will be the character that the brain is transplanted into.
  • An amiibo .bin file that has also been formatted for Super Smash Bros.
  • Tagmo, version 2.6.1
  • A backup of the amiibo being overwritten, if you want to reverse the transplant

 

Step 1:

Turn on the NFC feature on your Android phone. Start the Tagmo app. Check the box labeled “Allow restore to different tag”.

(You should learn beforehand where the NFC writer is located on your phone, so as to avoid corrupting your amiibo hardware)

Step 2:

Press “Load Tag”, and select the tag of the amiibo “brain” you wish to transplant into the amiibo hardware. If you wish to place an amiibo into a Ganondorf’s body, you MUST have a Ganondorf figure, NTAG215 chip or card, or a Powertag that has been formatted to be a Ganondorf amiibo.

Step 3:

Press “Restore Tag”, and then place your amiibo hardware on the NFC sensor on your phone. It may take a few tries to properly transplant the .bin file onto your physical hardware.

Step 4:

Scan your amiibo into Super Smash Bros. and confirm that the transplant worked. If the transplant succeeded, then your amiibo hardware will have a different name on the menu screen- it will be the name that the .bin file amiibo had. The amiibo hardware will also have all rewards,  moves and stat loadouts that the .bin file had. The only thing that wasn’t transplanted was the character of the .bin file.

What does this do?

This takes all the training and experience that the .bin file amiibo had acquired and places it into the hardware. In a sense, you are transplanting the .bin file brain into the body of another, different character.

This is a pure transplant- if you take a Ness who has been trained to use his Side-B and back throw move, and place him into a Ganondorf amiibo, it will be a Ganondorf amiibo that uses his Side-B and back throw. Absolutely everything about your amiibo will be the same except for their character.

We Don’t Know How Much About Amiibo that We Don’t Know- But it’s Probably a Lot

If you’ve spent a bit of time hanging around the last corners of the web in which amiibo is still a topic of discussion, you’ll notice an all too common problem for online communities. There’s a select group of people who are absolutely certain their opinions are correct. And to an extent, they have a reason to believe that: after all, amiibo training and competitions have been going on since the release of Super Smash Bros for Nintendo 3DS and Wii U. Surely after three years we would collectively have a pretty solid grasp on the abilities of amiibo, right? So it’s not unreasonable for someone to assume their knowledge about $13 plastic toys is complete.

Except that nobody’s knowledge about amiibo training is complete, or anywhere near complete. After all these years and all these tournaments, the tier list is still shifting around. New loadouts are being tested for previously unviable amiibo, and new training methods are being developed. Tiny nuances have been discovered as well: the head honcho of Cloud Nine, the other amiibo website, just found out that the Improved Launch Ability bonus significantly improves the damage output of the amiibo assigned to it.

In fact, a MASSIVE discovery was made last week for amiibo training. Someone figured out (still working on finding out who), using a defunct version of the Android app Tagmo, how to take the training that an amiibo has received and put it in a different character. If I understand correctly, this means that you will basically have the same amiibo in a different body. By taking a Shulk who uses his Down Special all the time and putting him in a Bowser, you would get a Bowser who uses his Down Special all the time.

While the amiibo community is still tinkering and researching the possible combinations that would be beneficial (well, part of the amiibo community; it seems the rest of them are complaining that this discovery is “overhyped”), one thing is for certain. We don’t know the limits of this new method of amiibo training. It could be the future. It could be useless. We don’t know how much about amiibo that we don’t know- but it’s probably a lot.

Amiibo Science: Bone and Flesh, the innate structures of amiibo

Get out your degrees, because today the Amiibo Doctor is going to explain an important amiibo concept! We’re going to discuss, using the skeletal structure as a metaphor, how amiibo intelligence is structured, and how that structure allows stimuli to change its behavior.

If you paid attention to my post from yesterday, you’ll remember that amiibo of the same character end up using the same moves over and over again, regardless of their utility. Yoshi’s egg throwing was a useful example of this: if you train Yoshi while playing as Yoshi and NEVER use his egg throw, your amiibo will still eventually use the egg throw and only the egg throw.

Most amiibo have a tendency like this, but it’s not as obvious which move or moves they prefer. It may be obvious that Sheik has an obsession with her jab, but it’s not as obvious that a Mario will end up using his neutral air and his down smash, given enough time. Sometimes you have to really look for these inherent favorites: I trained five individual Mario amiibo before I realized that after a hundred matches, they all ended up doing the same thing.

(If you’re a trainer who wants to prevent their amiibo from using moves, we’ll discuss that in a later post.)

However, amiibo never perform an action 100% of the time, and thus will change over time, even if it is only in slight ways. For example, if you pit your amiibo against an opponent for ten matches, you had better believe that it’s going to come out different than it went in. At the very least it’ll learn a bit better what not to do.

 

Now let’s say you really pay attention, and you pit your amiibo against an opponent for ten matches ten times, for a total of one hundred matches. You’ll notice something unusual for artificial intelligences: some things about the amiibo change, but some don’t. Perhaps it still wants to use its down smash just like it always has, but now it has switched out using one move for another. If you spot this, know that it is the same situation as Yoshi, but it just isn’t as innate. Your amiibo is not about to use only one move for the rest of its life. Don’t worry. In fact, take note of the moves that it’s using: they are the bone, if you will. They are the structure that everything else is built from.

But a bone is useless by itself. It can’t do anything without flesh on top of it to move the bone. And just as with a bone, an amiibo needs more than just its inherent move preferences to fight well. So we have to put on some flesh.

In the metaphor, the “flesh” are the experiences that the amiibo has obtained since it was first taken out of the box, or more specifically, the changes made to amiibo’s hex data since it was originally formatted for Smash Bros. The bone is always the same: the flesh grows and changes shape to adapt to its environment. Just as lifting weights to make your arm larger will change the muscle patterns in your arm, teaching your amiibo to do a certain thing will change the data of the amiibo.

In summary, it is the experiences, also called the “flesh” that makes each amiibo unique. And the flesh is always built off the same bone.