Math Suggests 45 Characters (+x hidden characters) to be Playable

[erico9001]

I used a spreadsheet to try to see how many characters would be in Smash bros 4. As you can see, I found the announcement date for each character, compiled them to create the amount of days from e3 2013 and compared it to the amount of revealed characters since e3 2013.

I'm quite pleased with this, actually. R² is really close to 1. It also turns out I was correct back when the game was first announced in thinking that the frequency of smash characters would increase with time, but this is irrelevant.

So anyways, there are 460 days between e3 2013 and the release date for Smash bros in Japan. It's pretty obvious that by that date we would have all the characters announced except for the possible secret ones. Extrapolate the trendline out to x=460 days, and you get y=44.83 characters. Round to 45, unless we're looking at 8/10 of a character! (I guess that could be Little Mac =P). Add in your preference for the amount of secret characters. Personally, I see it being 3 only based on how I think there was supposed to be 2 back in Brawl before they got leaked or something. I think I read this somewhere a while ago, but idk, it's just a guess.

The equation is 11.7522786583exp(.0029106077x) so that if you want to test it out yourself you may do so without having to type it out. (google can do this if you just paste it in)

Here's a download for the spreadsheet, but I don't think the graph will show up.
https://dl.dropboxusercontent.com/u/275120798/Smash bros characters.xls

And here's my data for the characters, when they got revealed, and the amount of days between their announcement dates. The characters fall under the date they got announced and then the time elapsed between the next character announcement is below that.

Spoiler

June 11

Mario, Donkey Kong, Link, Samus, Kirby, Fox, Pikachu, Bowser, Pit, Villager, Mega Man, Wii Fit Trainer

31 days

July 12

Olimar

26 days

August 7

Luigi

36 days

September 12

Peach

14 days

September 26

Toon Link

5 days

October 1

Sonic

37 days

November 7

Marth

31 days

December 18

Rosalina

8 days

December 26

Zelda

15 days

January 10

King Dedede

21 days

January 31

Lucario

13 days

February 13

Little Mac

8 days

February 21

Diddy Kong

46 days

April 8

Zero Suit Samus, Sheik, Yoshi, Charizard, Greninja

45 days

May 23

Ike

18 days

June 10

Miis, Palutena, Pac-Man

34 days

July 14

Captain Falcon, Robin, Lucina

... and that's what the math suggests! It's nothing complex, really. I've double checked all of the data, but still, if anyone find a mistake feel free to let me know and I can make adjustments. As more characters get announced, I can update this, but it probably won't change much unless troll Sakurai REALLY throws a curveball.

Feel free to use the data I gathered if you want to try doing it a different way. It took a while to find all of that.

 

[SKM_NeoN]

Math stinks man, keep that witchcraft far away from me!

And your avatar scares me.

Edit: Looks like you changed your frightening avatar, lol!

 

[Thedude3445]

There were 11 secret characters in Melee, and 12 in Brawl (if you ignore Sonic and Snake). Since Luigi is possibly the first unlockable revealed, and Sonic, Pac-Man, and Megaman will probably also be unlockable, I'll go ahead and say there will be about ten.

With this math that makes 55 characters and that's absolutely insane.

 

[shapular]

The thing with regressions is that they're really only good for predicting data inside the range of the data you already have. Trying to predict a value two months past your latest point with an exponential regression probably isn't very useful.

 

[Phaazoid]

This is called extrapolating. It's using previous trends to try to predict data outside of the current bounds of our information. And real mathers know that results from it can easily be completely false.

Don't be like Fox News. Don't put your faith in extrapolation, kids.

 

[MasterOfKnees]

We have 36 characters currently, so let's add the shoe-ins to fit 45: Wario, Ice Climbers, Meta Knight, Mr. Game & Watch, R.O.B, Falco, Ganondorf, Ness and Jigglypuff

No more newcomers and no Lucas or Wolf. Somehow I find that very unlikely.

As for secret characters, now that we have 4 FE character revealed I think that has more or less been thrown out the window.

 

[Ryuutakeshi]

That curve could easily shoot up exponentially. It's a decent idea, but you've already admitted it doesn't take into account "secret characters" so it doesn't even give us a guess at total roster size.

Also, relevant:

Spoiler

 

[erico9001]

There were 11 secret characters in Melee, and 12 in Brawl (if you ignore Sonic and Snake). Since Luigi is possibly the first unlockable revealed, and Sonic, Pac-Man, and Megaman will probably also be unlockable, I'll go ahead and say there will be about ten.

With this math that makes 55 characters and that's absolutely insane.

By secret characters I'm referring to characters that weren't announced to be in the game prior to the release date, not unlockable characters.

I believe that it was Jigglypuff and Wolf that were planned to be hidden to players until they got revealed in game, that is, until a leaker revealed them.

The thing with regressions is that they're really only good for predicting data inside the range of the data you already have. Trying to predict a value two months past your latest point with an exponential regression probably isn't very useful.

60 days isn't very much of an extrapolation since we've already had 388 days of observation.

This is called extrapolating. It's using previous trends to try to predict data outside of the current bounds of our information. And real mathers know that results from it can easily be completely false.

Don't be like Fox News. Don't put your faith in extrapolation, kids.

All of science is based on extrapolating data. It's how any prediction is made. If we want to have a reliable prediction, meaning one with no opinions skewing the results, well... in the words of Farnsworth when he's in Bender's body..

Faith? Wtf. Do you even know what faith is? This is pretty much the opposite

We have 36 characters currently, so let's add the shoe-ins to fit 45: Wario, Ice Climbers, Meta Knight, Mr. Game & Watch, R.O.B, Falco, Ganondorf, Ness and Jigglypuff

No more newcomers and no Lucas or Wolf. Somehow I find that very unlikely.

As for secret characters, now that we have 4 FE character revealed I think that has more or less been thrown out the window.

Please remind me, how are Rob and G&W shoe ins?
Anyways, my subjective number of hidden/secret characters is 3, so I look at it more like 48. I'd consider it possible for Lucas, Rob, and G&W to be cut, so that leaves it at 44 with the possibility of 4 more newcomers to appear. That's a lot. There's room for like Shulk, Ridley, K.Rool, and idk, chorus kids or something to fit into there. Two of those I don't even see as likely at all. Perhaps I'd get rid of two and put back in Lucas and G&W.

The thing is, how many characters are you people really expecting to be announced in these next two months? I mean, my graph is already predicting 9 freaking characters to be announced in just 2 months. There was just a reveal, too, so you have to take that into account as well.

That curve could easily shoot up exponentially. It's a decent idea, but you've already admitted it doesn't take into account "secret characters" so it doesn't even give us a guess at total roster size.

Also, relevant:

Spoiler

The upward curve makes it so by the end we're getting a new character every like 8 days. It's already pretty damn high near the end. As for the picture, there are 17 points of data and the trendline matches really closely with them. I don't even see the purpose in posting such an image.

-

*sigh* guys, take from it what you want. Yes, it's possible for it to be off by one, or even, hell, even two characters, but the point is that this is a prediction. Math 'suggests' 45 +x secret characters, not Math 'states' this. If you're looking for an objective base line to your predictions on how many characters will be in Smash 4, this would be that. Then, go ahead and modify it to your hearts content with subjectivity and opinions. You can even make the amount of hidden/secret characters (things not revealed before release) any number you want! I'm a realist, and I went with the realistic number of 3, but you can go ahead and say there will be 30 if you so want. If you are working things out and you find yourself one short of the character you want to be in the game, go ahead and assume that the extrapolation is off by one. It very well may be.

 

[leprechaunlink727]

We've already got 36 announced characters, so let's factor in the characters that are definitely coming back no matter what: Ness, Jigglypuff, Wario, Meta Knight, Ice Climbers, Ganondorf, Game & Watch, Falco, and yes, ROB. That brings us up to 45 already. We already know there's at least one newcomer left that we haven't seen, but I'm inclined to believe there's actually between 3 and 5. Then there's an additional four vets who could still return: Lucas, Wolf, Snake, and Mewtwo. So 48 at the absolute lowest, and 54 at best.

 

[NoiseHERO]

Isn't Sakurai the type of dude to flip the table and reveal 10 people in one trailer on a thursday afternoon with half of them being characters no one thought of though.

 

[TumblrFamous]

How are ROB and G&W likely cuts. How is that possible.

 

[TerminalMontage]

How are ROB and G&W likely cuts. How is that possible.

G&W probably passed the torch to Pacman, why would they show a character, but not officially announce them? G&W is probably not in.

 

[MasterOfKnees]

Please remind me, how are Rob and G&W shoe ins?
Anyways, my subjective number of hidden/secret characters is 3, so I look at it more like 48. I'd consider it possible for Lucas, Rob, and G&W to be cut, so that leaves it at 44 with the possibility of 4 more newcomers to appear. That's a lot. There's room for like Shulk, Ridley, K.Rool, and idk, chorus kids or something to fit into there. Two of those I don't even see as likely at all. Perhaps I'd get rid of two and put back in Lucas and G&W.

Game & Watch was practically confirmed in Pac's trailer, he even has a new look. Why he's not on the site I can't say, but he's definitely not out. R.O.B I just don't see why would be cut, I'd definitely consider him of higher priority than guys like Lucario, Ike and Toon Link, all who have returned. 3 hidden characters also seems odd, why not just reveal everyone if you're so close anyways?

We're going to see an upwards spike in character reveals these last weeks I'm sure, we even had 3 character confirmations in a single trailer just yesterday. Math and curves are very unreliable in this scenario.

G&W probably passed the torch to Pacman, why would they show a character, but not officially announce them? G&W is probably not in.

And Mario passed the torch to Donkey Kong I guess? Seriously, G&W went over to Pac-Man and rang his damn bell, how does one interpret that as passing on the torch? And what torch? The 1980 torch? That sounds beyond bizarre.

 

[TumblrFamous]

G&W probably passed the torch to Pacman, why would they show a character, but not officially announce them? G&W is probably not in.

Because they were simply referencing the first video game characters. If we didn't get a confirmation of Captain Falcon in yesterday's reveal, would he be cut then?

And how does he even "pass the torch"? The only thing similar is that they're early characters.

 

[TerminalMontage]

Because they were simply referencing the first video game characters. If we didn't get a confirmation of Captain Falcon in yesterday's reveal, would he be cut then?

And how does he even "pass the torch"? The only thing similar is that they're early characters.

If they showed Captain Falcon in a video, but didn't show gameplay of him, and didn't show him on the site.. I'm willing to bet that means he isn't on the roster. I don't see why they would show G&W and not put him on the site. Plus, his animations look weird, as if they were thrown together for the sake of this video, it wasn't like his Brawl model or anything, he looked different.

He and Pacman even share similar moves, at least it looks like they do. Pacman's down smash seems to operate like G&W's, then he has that trampoline jump thing, and maybe a few others. Although, I hope I'm wrong. G&W is my main.

 

[NoiseHERO]

I do find it weird that this late in development, they would show G&W in any instance (not even sure how long it takes them to make these trailers as well.) and not show actual gameplay footage, I mean he's already in the trailer in general....

But I also think it's kind of messed up to show in AT ALL if he's not in the game.

 

[leprechaunlink727]

Guys, Game and Watch is going to be playable. I will shave my head and post pictures if he's not.

 

[erico9001]

How are ROB and G&W likely cuts. How is that possible.

I never said likely cuts. They're possible cuts. I hold subjective opinions and ideals despite the objectivity of the graph. I chose G&W and ROB as examples in my reply because I felt more people would be inclined to agree with my on them, but really because I was apalled that he believed 100% they were in. Not even a 1% chance they weren't. Whether they are in or not isn't really the point here, though.

Adding up the various characters and deciding who gets in based off of that is based on preference. However, the graph makes a prediction based off of fact. We've made it to the point where we have enough data to make this prediction using the facts instead of our preference. There is room for error, as I said, since the graph could be off by a character and we still have no way of telling how many hidden characters there will be. Keep in mind, however, that the graph could also be off by one too many, not necessarily by one too little.

So pretty much, when you're making your character prediction and you come across a number of 55, perhaps consider that that number is too high. If I may reiterate something I said in my previous post, 9 characters in two months is already a lot compared to what we've been getting. The rate is accelerating, and the graph shows and accounts for that. When the next character reveal comes, this will only be more accurate.

 

[TumblrFamous]

I never said likely cuts. They're possible cuts. I hold subjective opinions and ideals despite the objectivity of the graph. I chose G&W and ROB as examples in my reply because I felt more people would be inclined to agree with my on them, but really because I was apalled that he believed 100% they were in. Not even a 1% chance they weren't. Whether they are in or not isn't really the point here, though.

Adding up the various characters and deciding who gets in based off of that is based on preference. However, the graph makes a prediction based off of fact. We've made it to the point where we have enough data to make this prediction using the facts instead of our preference. There is room for error, as I said, since the graph could be off by a character and we still have no way of telling how many hidden characters there will be. Keep in mind, however, that the graph could also be off by one too many, not necessarily by one too little.

So pretty much, when you're making your character prediction and you come across a number of 55, perhaps consider that that number is too high. If I may reiterate something I said in my previous post, 9 characters in two months is already a lot compared to what we've been getting. The rate is accelerating, and the graph shows and accounts for that. When the next character reveal comes, this will only be more accurate.

As others have said, I just don't think you should fully rely on quantitative measures to predict character numbers and reveals. We've seen how unpredictable speculation has been.

 

[Phaazoid]

All of science is based on extrapolating data. It's how any prediction is made.

This is very, very wrong. and false. I appreciate your futurama reference, but there is a major lack of understanding of science going on here. Science is based off of the scientific method, which is making a hypothesis, testing the hypothesis, and making judgements based on the results of that test. Not extrapolating points on a graph. Ideas are a different beast than numbers altogether. You have confused induction with extrapolation. Induction is... sometimes a hard thing to really understand. Mathematical induction exists, and it too can be confused with extrapolation.

Basically the takeaway is

Induction =/= Extrapolation

 

[SKM_NeoN]

You guys are looking at this the wrong way. This isn't a prediction, it's a guesstimate based solely on the sporadic pattern of reveals we've been given from a pure mathematical standpoint, and it has its merits. In fact, this number might be much closer to the actual roster size than a lot of you seem to think. Here's an interesting quote from Mr. Sakurai:

“It isn’t a matter of ‘if the next game has 50 characters, that’ll be enough. There is a certain charm to games that have huge casts of playable characters, but they tend to have issues with game balance and it becomes very difficult to fine-tune each character and have them all feel distinctive…. In terms of quantity, we’ve probably already reached the limit of what’s feasible. I think a change of direction may be what’s needed.”

When you take all the research into account, I think @ erico9001 might be on to something here.

 

[smashingDoug]

We have 36 characters currently, so let's add the shoe-ins to fit 45: Wario, Ice Climbers, Meta Knight, Mr. Game & Watch, R.O.B, Falco, Ganondorf, Ness and Jigglypuff

No more newcomers and no Lucas or Wolf. I find that very unlikely.

As for secret characters, now that we have 4 FE character revealed I think that has more or less been thrown out the window.

fixed for you buddy

 

[erico9001]

This is very, very wrong. and false. I appreciate your futurama reference, but there is a major lack of understanding of science going on here. Science is based off of the scientific method, which is making a hypothesis, testing the hypothesis, and making judgements based on the results of that test. Not extrapolating points on a graph. Ideas are a different beast than numbers altogether. You have confused induction with extrapolation. Induction is... sometimes a hard thing to really understand. Mathematical induction exists, and it too can be confused with extrapolation.

Basically the takeaway is

Induction =/= Extrapolation

I don't really want this to be the main discussion so I'm spoilering it:

Spoiler

In order to make any sort of prediction we must establish patterns between what we know and then use those patterns to predict other events. While we all may not be doing this with excel, we are at least doing it unconsciously inside our heads. This is what I mean. In Geology, scientists search for new points of data, whether they be fossils or geological formations, so that they can improve the pattern we are creating as far as the functions of Earth go. We can't simply dig up physical evidence for all the answers in a field though. There are just fragments. Through the patterns of the fragments is when we can get the real of idea of what is going on. These patterns are applied not just to the past in order to figure out what did happen, but also to the future as to what will happen. We also find Geology's findings being applied into Astronomy as it turns out we can connect our findings on Earth to other planets. On the other hand, we have theoretical physicists who focus on discovering the deepest, most universal patterns... or, using the patterns we have already established in order to gain more information/data, to just be used again for more patterns.

You guys are looking at this the wrong way. This isn't a prediction, it's a guesstimate based solely on the sporadic pattern of reveals we've been given from a pure mathematical standpoint, and it has its merits. In fact, this number might be much closer to the actual roster size than a lot of you seem to think. Here's an interesting quote from Mr. Sakurai:

“It isn’t a matter of ‘if the next game has 50 characters, that’ll be enough. There is a certain charm to games that have huge casts of playable characters, but they tend to have issues with game balance and it becomes very difficult to fine-tune each character and have them all feel distinctive…. In terms of quantity, we’ve probably already reached the limit of what’s feasible. I think a change of direction may be what’s needed.”

When you take all the research into account, I think @ erico9001 might be on to something here.

Yeah, pretty much actually . I also think it's important to pay attention to what Sakurai said there about the size of the roster.

 

[Number2joker]

we need a grass starter from pokemon!

 

[DakotaBonez]

Not even math can calculate what Sakurai's got planned.

 

[GamerGuy09]

Credit to user Bmaick on Reddit:

The guy posted this on reddit as how the Menu may look like will all of the newcomers + brawl vets - Squirtle Snake and Ivysaur.

I can say with certainty that there will be at least 4 more characters, as Sakurai said Robin is not our last Newcomer obviously.

So I'm pretty sure a roster of 51 is nearly guaranteed.

 

[andimidna]

There were 11 secret characters in Melee, and 12 in Brawl (if you ignore Sonic and Snake). Since Luigi is possibly the first unlockable revealed, and Sonic, Pac-Man, and Megaman will probably also be unlockable, I'll go ahead and say there will be about ten.

With this math that makes 55 characters and that's absolutely insane.

That's not insane as long as clones are involved.

Think about who some of those unlockables were..
And now think about Lucina... Is she really going to be the only one? I mean, Lucina would have been my first pick for a clone but that's just bias. Sakurai has already referred to Lucina's situation as if it wasn't unique to her. Something like "when characters differentiate slightly from who they're based off of they get their own slot on the roster"
55 is now plausible because of this. And also fits with what seems to work with the demo roster size (47,51,55)

 

[bilbo43]

Ive held the belief that its going to be around 30 days before nintendo officially reveals their next batch of newcomers, and who knows how many gets revealed there or who. We also dont know which veterans will be revealed alongside them.

Just as a personal hunch, I'm expecting a villain to appear in the next video, either ganondorf fighting someone like shulk or someone completely new.

Also dont forget that there is supposed to be a direct during august (not 100% sure on this, just seems to be a generally thought on the boards, despite no official date that im aware of) and apparently corocoro will be revealing the 3DS lineup (starter or full is yet to be known). So we will get a fair share of characters in my honest opinion within the next 30 days.

Just because we get 3-5 characters at the direct and if only the starter roster is revealed in corocoro then we still have the potential for secret characters to appear in the actual game.

I do get the feeling however that all the starting roster has been revealed, with the exception of maybe some obvious inclusions like Ice Climbers and Wario. So maybe the corocoro scans wont actually reveal (many) new characters and only give us who we can expect to start off with.

 

[DakotaBonez]

Feel free to use the data I gathered if you want to try doing it a different way. It took a while to find all of that.

Did you count the miis as 3 characters, if not, could ya change the data so that the miis count as 3 characters?

And to think, a couple years ago when we were speculating roster size, everyone was saying 50 was WAAAAY outta reach.
We've already got 38 confirmed characters ( counting miis as 3 characters)
And there's still at least 2 newcomers to be revealed (40)
And if ya add in all of the unconfirmed veterans (Ness, Falco, Ganon and Metaknight to name a few)
We could pass 50 easily. (well maybe not that easily)

To be safe I'm gonna guess 47, but if ya count the mii fighters as seperate than you've got 49. But who stops at 49 when you're one away from 50!?!

 

[DakotaBonez]

Double post...

 

[erico9001]

Did you count the miis as 3 characters, if not, could ya change the data so that the miis count as 3 characters?

And to think, a couple years ago when we were speculating roster size, everyone was saying 50 was WAAAAY outta reach.
We've already got 38 confirmed characters ( counting miis as 3 characters)
And there's still at least 2 newcomers to be revealed (40)
And if ya add in all of the unconfirmed veterans (Ness, Falco, Ganon and Metaknight to name a few)
We could pass 50 easily. (well maybe not that easily)

To be safe I'm gonna guess 47, but if ya count the mii fighters as seperate than you've got 49. But who stops at 49 when you're one away from 50!?!

Hmm. I'm going to keep the 'mii fighter' one character because on the official site it shows up as just one. If you're curious though, I can adjust the data just to satisfy curiosity about what if they were 3 characters.

Ok, so if the miis are 3 characters instead of 1... the equation becomes 11.6156820724exp(0.0030071805x), the R² (coefficient of determination) is lowered a tiny bit to .988, and at x=460 days, y= 46.33 characters.

Just for fun, I adjusted the original for if Lucina isn't a character and got the equation 11.7917954279exp(0.002884171x), R² of .993, and 44.44 characters. It's .9 characters less than if she was a character, so t's pretty much just like subtracting her anyways.

 

[erico9001]

Actually, I think 49 characters may be more likely because it makes an even 10 colums × 5 rows (with random)

 

[Banjo-Kazooie]

Yeah sure, I'm willing to believe any crazy theory you have that gives me more characters.

 

[AlvisCPU]

At the risk of repeating others, I'm surprised people are treating this more as a truth. It's not claiming to be confirmation of anything. Yes, it's extrapolation so we need to be careful, but long as we're all aware of what considerations and assumptions were taken in making it, and we treat it with validity that reflects those considerations/assumptions, it's okay.

I personally think this is a pretty good indicator of what we'll see, an R² that high is hard to argue against. It gets pretty steep towards the end so that's when we'll have to get a bit hesitant with accuracy, seeing maybe ~3 more character reveals in the coming months isn't really a stretch.

Wish I'd had the idea to do this earlier on, nice work!

 

[pupNapoleon]

I will investigate this later. This pleases me much more than the ridiculous and poorly founded 47 for CSS case, which took into account an all of 'the character sizes cannot shrink.'

Still plenty of variables could be at play here, but lovely going!

 

[MetroidFan4Life]

The more characters they add the better :3