Tournament Videos from Japan

[AlexAnthonyD]

Thank you so much for posting these I really loved Nietono's gameplay. I can only imagine two Greninja's going at it in a style like he plays and then I will really laugh at anyone who called this game slow paced. In fact he made Greninja seem like the perfect counter or to any campy player if that type of match aggravates you.

 

[Fafnir]

Essentially, in a game that uses an RPS style of "balance," it quickly ends up degenerating into an uncompetitive environment where entire sets can be decided at blind pick. It's not at all healthy for any game.
While perfectly balancing every aspect of every match up is essentially going to be impossible so long as each character has measurably different traits, the goal is to bring that gap as close as possible. Probably the best way to start with this is to give every character the basic tools needed to get themselves out of any specific situation that another character might put them into. Granted, the tools for a specific purpose on one character might be severely lacking in capability to those of another, but they should exist in some way, shape, or form.

And I'll have to agree with Thinkaman on the subject of Extra Credit's "Perfect Imbalance" episode. It was an utterly horrendous episode of a series that I otherwise typically hold in high esteem, and all the worse because it always ends up being brought up in any balance discussion for any game.

 

[petrie911]

This depends exclusively on your fitness function.

The standard deviation of matchup ratios is lower in the second part.

I would much rather have two 55/45s than one 60/40, just as I would rather have ten 55/45s than one 100/0.

This is true for both sides of the coin, winner or loser.

If you exclusively care about characters being picked equal amounts (in and of itself), then the former is superior.

But if you have any other criteria at all, such as wanting the game to be as balanced as possible, the latter is superior.

It's not so much characters being picked in equal amounts as it is about not having characters who aren't picked at all. Having those 55-45 matchups instead of 60-40 ones is pointless if no one is playing them. This happened to Starcraft 2 in its early days. Terran was advantaged against Zerg and Protoss for quite some time. The advantage was only about 55-45, but it led to the GSL code S being 31 Terrans and 1 Zerg. There is no world in which that can be considered well-balanced.

 

[Thinkaman]

It's not so much characters being picked in equal amounts as it is about not having characters who aren't picked at all. Having those 55-45 matchups instead of 60-40 ones is pointless if no one is playing them. This happened to Starcraft 2 in its early days. Terran was advantaged against Zerg and Protoss for quite some time. The advantage was only about 55-45, but it led to the GSL code S being 31 Terrans and 1 Zerg. There is no world in which that can be considered well-balanced.

But there is a critical point where changing characters and going against one's skill affinities and playstyles is a steeper cost than the matchup differential.

Jigglypuff was a pretty poor character in Brawl, definitely bottom 20%, maybe lower. But Jigglypuff was still my best character since I am way better at aerial spacing than any other skill test in the game. The affinity gap for me was less than the character's gap in matchups.

For a lower gap in matchups, this effect scales exponentially.

This is why there were so few Meta Knight players in Brawl: less than 10%, rather than 100%.

Again, it all depends on your criteria:

• If all you care about is characters showing up and having their names on a list, counters will help.
• If you actually care about the fairness of the gameplay in each individual possible match, counters that determine or bias the winner from character select are very bad.

For example, Pikachu was not a very common character in Brawl. (Pretty solid, but oddly rare.) The fact that he had an infinite on Fox, a modestly strong character, increased the amount of Pikachu play non-trivially.

This was "good" for on-paper diversity (number of characters used in a tourney), but awful for the actual gameplay.

An infinite causing a 90/10 matchup is an extreme case, but it is just as flawed (just lesser magnitude) to have even a 60/40 counter.

 

[Fafnir]

It's worth noting that just about any meta will end up becoming centralized at some perceived top (even Pokemon, which runs a tier list based on how centralized specific pokemon make the game, ends up with centralization within its tiers). Every fighting game is going to have its best character, even if that character is only the best by a small margin. The goal is to make every character as viable as possible against one another so that they can all compete against one another. At the end of the day, the goal should be to play against your opponent, not against the match-up.

One major problem with the concept of RPS style balance (and there are a lot of huge problems) is that should a character become the eventual top tier, every single match up that loses to them becomes unviable. We saw this with every single character in Melee that got chaingrabbed by Sheik. While there might be other top tier choices that end up losing to such fringe match-ups, the characters that express such dominance are so strong in their position that these "counters" really don't matter. The RPS meta ends up persisting in just as unhealthy a way as ever, but with a much smaller and shallower metagame. That's how you end up with nothing but Fox/Falco>Sheik>Marth>Fox/Falco.

 

[Raijinken]

It's worth noting that just about any meta will end up becoming centralized at some perceived top (even Pokemon, which runs a tier list based on how centralized specific pokemon make the game, ends up with centralization within its tiers). Every fighting game is going to have its best character, even if that character is only the best by a small margin. The goal is to make every character as viable as possible against one another so that they can all compete against one another. At the end of the day, the goal should be to play against your opponent, not against the match-up.

One major problem with the concept of RPS style balance (and there are a lot of huge problems) is that should a character become the eventual top tier, every single match up that loses to them becomes unviable. We saw this with every single character in Melee that got chaingrabbed by Sheik. While there might be other top tier choices that end up losing to such fringe match-ups, the characters that express such dominance are so strong in their position that these "counters" really don't matter. The RPS meta ends up persisting in just as unhealthy a way as ever, but with a much smaller and shallower metagame. That's how you end up with nothing but Fox/Falco>Sheik>Marth>Fox/Falco.

Going along with this, what normally happens with this is you get a slightly cyclical metagame with players trying to counter the "top" characters. Eventually, players pick whatever is the safest bet and/or counters the most, and unless that has a very strong counter which is also viable, that safest bet becomes the top character.

In Dota, the best strategy for the past three months was to get a strong pushing lineup and just snowball down the lanes, accumulating a tower gold advantage that was hard to surpass. Any theoretical counter-heroes to the popular choices like Lycan, Death Prophet, and Shadow Shaman were offset by the sheer effectiveness of pushing power. Even in the pro scene where you have to go through a picking and banning phase to get your team comp (as the game is designed with specific roles and numerous fairly hard counters in mind), there were still many strong pushing lineups, and while banning the top heroes for the meta occasionally made the game less stale to watch competitively, the strategy was the same.

Thankfully, Dota gets patches after a while. Most fighting games don't, though there's always the hope/chance that Smash will in this newest version. Of course, in Smash, we strongly prefer for the game to be based on player interactions and not so much character interactions, but at the same time, it is basically a given that no two characters of different playstyles (take Duck Hunt and Little Mac for example) will ever be perfectly matched.

A bit more on topic, though: Those rounds were great, and have me very closely considering Pac Man as a main now.

 

[MugenLord]

Just like I said before and hopefully I am right. This game looks more like it plays more like a traditional fighter than any other smash. It seems as though that the depth is found more in the characters themselves than advance controls and movements. Then it looks as if you will need to have more than one character learned in order to counter match-ups and that is a good thing. I am really digging this game, almost the entire cast plays differently and has something different to bring to the table.

 

[~ Gheb ~]

Small Update of the OP with videos of a recent Wifi Tournament including top level play for Diddy Kong, Sheik, Peach and Rosalina. Make sure not to miss these videos!

 

[mangamusicfan]

Good matches thanks for the link.
pacman cam be a real beast!

 

[Noa.]

Diddy is so definitely top tier. He has so many options and so much control over the match. Shame that the Rosalina player SD'd so much.