High-Reward Characters

[ibrium]

For those who aren't going to read the following paragraph, this basically lists characters who have a high reward for learning them inside out and backwards. Thanks, XienZo >>;

Please do not take this as an attack on established tiers or match-ups or anything like that. It's simply an interesting look at the data on tournaments. If you don't think it means anything, that's fine. I do- Just not necessarily exactly what it says.

At any rate, the following information is something I found while scrutinizing the wonderful list created by Ankoku. I'm sure we're all very much aware of the way he collects his data, and I'm very happy to have such reliable information on hand.

Using his collected information, I created a sort of "secondary list" that doesn't exactly tell the same story. While the ranking list accounts for all sorts of variables (such as tourney size, etc), this one does not. At least not yet.

The following are numbers I've currently evaluated using 2 simple inputs: One, the TOTAL PLACEMENTS reported by Ankoku, and two, the total wins reported by the very same. I used them to create what you might call a win percent: that is, if the character were to place in the top 8, how likely it would be that that character would win the tournament.

It is interesting because it reveals how characters tend to do at higher levels of play. I make the assumption that, by making it into the top 8, the player controlling the character must be marginally better than those who did not. This helps to suggest that these character win %s hold true for higher levels of play-- though that, of course, depends on where the data came from.

The interesting result it yields is that if you want to spend your time getting "good" with one character, make it one that will win MORE OFTEN.

Now remember, I have no buffer for large/small tournaments, or for good/bad players at the moment. I'm simply reworking Ankoku's numbers to look at a slightly different angle. That being said, the character highest on the list may surprise you.

The %'s were determined as follows:

#wins /( # top 8 + # top 4 + # top 2 + # wins )

Pokemon Trainer

33%​

Pikachu

29%​

Lucario

26%​

Sonic

23%​

MetaKnight

22%​

Pit

20%​

Snake

19%​

Wario

18%​

Bowser

17%​

Luigi

17%​

Captain Falcon

17%​

Wolf

16%​

Falco

15%​

Lucas

14%​

King Dedede

13%​

Olimar

13%​

Fox

11%​

Zelda/Sheik

11%​

ROB

9%​

Marth

8%​

Diddy Kong

6%​

Peach

6%​

Ice Climbers

6%​

Mr. Game & Watch

5%​

Donkey Kong

3%​

Kirby

0%​

ZSS

0%​

Toon Link

0%​

Ike

0%​

Mario

0%​

Yoshi

0%​

Link

0%​

Ness

0%​

Jigglypuff

0%​

Samus

0%​

Gannondorf

0%​

*Please note that any character with 0% had NO WINS, even if they had many placements and/or are high on the ranking list.

** I didn't seem to make myself entirely clear. Again, this in NO WAY means that PT could beat MK or that G&W is low tier. All this says is that at higher levels of smash (those who would make it into the top 8), this is how likely it is that character would win.

 

[brinboy789]

lolwut. MK has 22% while PT has 33% ?????

 

[ibrium]

lolwut. MK has 22% while PT has 33% ?????

Yessir. Even though MK has the most wins, when compared to his PLACEMENTS, he only wins 22% of the time. I hope that makes sense.

On the other hand, of the top 8 placements PT has had, he has won 33% of them.

 

[XienZo]

I think you forgot that there might have been tons of PTs that didn't place at all.

So this shows what character will place the highest whenever they DID place, and only IF they placed.

So this would favor characters with an extremely high learning curve that does well when properly played.

Finally, it would also favor characters who have great pros to back them up. (Azen, Anther, M2K...)

 

[ph00tbag]

You're not counting the people that didn't place in top 8. This list fails.

 

[ibrium]

I think you forgot that there might have been tons of PTs that didn't place at all.

So this shows what character will place the highest whenever they DID place, and only IF they placed.

So this would favor characters with an extremely high learning curve that does well when properly played.

Uh. Yeah.

I didn't explain that fully?

 

[XienZo]

So then what does this list tell us?

Oh, ok, stopped being lazy and read the whole thing(sorry)

So this should tell us the highest learning curve with the best results?

Still, the pros can skew the thing, such as MK still being high when he should be low due to the massive amounts of placings he has, especially since a tournament with a MK winning can have 2 more MKs place, but...

 

[ibrium]

You're not counting the people that didn't place in top 8. This list fails.

That's kind of the point. :/

 

[ibrium]

The pro's definitely can skew it. So can tourney size.

I just thought it was interesting to look at. And it makes Metaknight seem less intimidating. o.o

 

[-Mars-]

This is one of the sorriest attempts at a thread I have seen in a long time.

 

[ibrium]

This is one of the sorriest attempts at a thread I have seen in a long time.

You'll have to excuse my brashness, but I don't see what's sorry about it.

 

[XienZo]

No one reads the top paragraph... I think you should change "likelyhood of winning" with "highest learning curve" or "high-reward characters"

 

[ibrium]

No one reads the top paragraph... I think you should change "likelyhood of winning" with "highest learning curve" or "high-reward characters"

Done and done.

 

[Col. Stauffenberg]

I don't see how this represents learning curve at all.

 

[Samsquamptch]

if the character were to place in the top 8, how likely it would be that that character would win the tournament.

This is something of a misstatement, your method does not take into account the fact that a character can place in the top 8 multiple times in the same tournament.

Consider the following data:

Tournament A
1.MK
2.MK
3.Snake
4.DDD
5.G&W
6.MK
7.Falco
8.Wario

Now metaknight's score would be 1/3, indicating a 33% chance that MK will win the tourney if he is in the top 8.
Clearly MK has a 100% chance to win the tournament.

Your formula shows something along the lines of 'if a player picks a certain character, and the player makes the top 8, what are the chances of that player winning the tournament with the aforementioned character.'

Realistically, as a matter of ranking characters in order of "reward for learning them inside out and backwards." it is rather ineffective as it takes into account essentially no variables or sources of error.

 

[ibrium]

I don't see how this represents learning curve at all.

o.o;

Either I'm really missing something, or just not saying something correctly.

The percents indicate how often someone who PLACES with a character WINS a tourney.

So it should tell us how often, when a "good player" uses a character, that character should win.

 

[Col. Stauffenberg]

If you could get the character data for ALL the entrants in the tournies, it might. Well, not really, but...

When you have only the top 8 for each, it's kinda pointless.

 

[ibrium]

This is something of a misstatement, your method does not take into account the fact that a character can place in the top 8 multiple times in the same tournament.

Consider the following data:

Tournament A
1.MK
2.MK
3.Snake
4.DDD
5.G&W
6.MK
7.Falco
8.Wario

Now metaknight's score would be 1/3, indicating a 33% chance that MK will win the tourney if he is in the top 8.
Clearly MK has a 100% chance to win the tournament.

Yes, it would be. I would assume that the ranking list includes tournaments where that is the case, and that is where I got my data.

Your formula shows something along the lines of 'if a player picks a certain character, and the player makes the top 8, what are the chances of that player winning the tournament with the aforementioned character.'

...Yes.

 

[ibrium]

If you could get the character data for ALL the entrants in the tournies, it might. Well, not really, but...

When you have only the top 8 for each, it's kinda pointless.

Should the top 8 not be the best?

I'd say that's kind of the point.

So this is how characters do when they are represented at their best.

 

[Col. Stauffenberg]

Ankoku's thread already shows how characters do when they are represented at their best.

 

[ibrium]

Ankoku's thread already shows how characters do when they are represented at their best.

I disagree.

His list reflects how often a MK places and wins and gives out points for how often.

This list shows how often a character would win if it made the top 8.

There is a difference.

His is affected by how many MK's play. Mine, in theory, is not. (Though obviously theory =/= practice)

 

[XienZo]

Basically, this thread searches to isolate how well they do, while Ankoku's is a compliation of other info, such as how many times they win, so its possible more interesting pieces of information could be missed unless isolated.

 

[Col. Stauffenberg]

Er, you're dividing it by the number of top 8 placings. It's very much affected by how many MKs play.

 

[ph00tbag]

This still determines nothing at all. If anything, it shows what characters are played almost entirely by players who place high anyway. It also doesn't take into account the weight of each win. That is to say, the entry fee and number of entrants have no bearing on the results. There is no rigor to these results. it's just random data that yields nothing interesting.

 

[ibrium]

Er, you're dividing it by the number of top 8 placings. It's very much affected by how many MKs play.

theory =/= practice

I guess what I meant to say is that it isn't neccesary to include every player's character because this list isn't concerned with every player.

Using the top 8 placements means that, if player A were to master a character to the point where he or she was able to make it to that level in the tournament, he or she would have such-and-such chance of winning, regardless of how many Metaknights or G&Ws there are.

This still determines nothing at all. If anything, it shows what characters are played almost entirely by players who place high anyway. It also doesn't take into account the weight of each win. That is to say, the entry fee and number of entrants have no bearing on the results. There is no rigor to these results. it's just random data that yields nothing interesting.

I continue to disagree. :/

More importantly, I already SAID it was a rough list becuase it had yet to account for such things.

The interesting result it yeilds is that if you want to spend your time getting "good" with one character, make it one that will win MORE OFTEN.

How is that not the case?

 

[Col. Stauffenberg]

Because Pokemon Trainer doesn't win more often. He just has incredibly poor placings and one of them happens to be a win.

 

[ibrium]

Because Pokemon Trainer doesn't win more often. He just has incredibly poor placings and one of them happens to be a win.

That's possible, yes.

Did I ever say "here, follow this as mantra?"

No. I don't think PT could ever be as good as MEtaknight.

But look how often G&W fails to place first compared to how many placings he has?

Or how often Pit does just the opposite?

that's what is interesting.

 

[ZeroFox]

PT has a higher percentage than MK? What is the world coming to?

 

[Col. Stauffenberg]

Pit and G&W have the same number of tournament wins in Ankoku's list. G&W has 5-6 times the placings in other areas. In light of that, it seems rather silly to say you're better off getting good with Pit than you are with G&W, doesn't it?

 

[Samuelson]

This fails more than John McCain.

Interesting idea but clearly gives misleading information.

 

[Pearl Floatzel]

This fails more than John McCain.

Interesting idea but clearly gives misleading information.

The oddest comparison I have seen this election.

But agreed. This could be interesting, but needs to have more thought and data put into it for actual results.
It's skewed because suppose MK's get top 8 in 100% of all tourneys, and wins 20% of all tourneys. (Not true) Suppose there are over 5,000 tourneys in the US in the times that have been recorded. (Also not true) If PT gets into top 8 in 3 tourneys ever and wins one ever, then he is rated better than MK, even though MK is in top 8 in all tourneys. Clearly MK will be more likely to win, but PT looks pretty good from that perspective.

Nice idea. Think it through and come back.

 

[ph00tbag]

Okay. Who won with Pokemon Trainer? Chuck Nasty. While also using Metaknight.

Lucario has Azen and Atomisk (who also uses a slew of other characters, including MK). Pikachu has Anther. Incidentally, Anther is also responsible for many of Sonic's "wins," and I guarantee you that for most of those, Pikachu played a greater role. The only other guy who won with him, DJ Browny, only went all Sonic in one tourney that had 14 entrants and a $9 entry fee.

That is why your results are meaningless. They fail, utterly and completely to take into account the reality of tournaments. When you take into account that a grand total of four or five of the best players in the country are responsible for this wins, often with the help of some of the more powerful characters in the game, your list utterly fails to represent anything meaningful.

 

[Tien2500]

I like any list with PT on the top.

 

[feardragon64]

Ok here is the problem with the list. You're assuming that the same number of people are putting the time into developing a character. You're not accounting for the fact that some characters are just played much more, even if not all of them place, it's more likely that there will be more placements. Counting the number of people in the top 8 only shows that there are generally x number of *character* mains that know their character really well and put the effort into developing them. That doesn't mean the characters themselves have a bigger payoff. If one of the people who main'd character A instead main'd character B, what's to say that character B wouldn't jump up in your percent table drastically? The number of competent players playing those characters has a lot more to do with the percents that you have than the actual characters.

tl;dr: Players affect this much more than the characters themselves.

 

[Genos]

** I didn't seem to make myself entirely clear. Again, this in NO WAY means that PT could beat MK or that G&W is low tier. All this says is that at higher levels of smash (those who would make it into the top 8), this is how likely it is that character would win.

Not true; getting into the top 8 in a tournament does not mean you are at the higher levels of smash. If you're at a tournament where everyone is terrible but you make top 8, that does not mean you are playing at the higher levels of smash. Rather, it proves that you are better than most people there. In this case everyone was bad, so your high placement means little.

Also take into account that tournament wins with bad characters are much more common at tournaments where everyone is bad. There are less limitations at these types of tournies since no one is playing their character to the best of of their character's ability. Of course a Pokemon Trainer could win if no one is able to abuse his weaknesses

Okay. Who won with Pokemon Trainer? Chuck Nasty. While also using Metaknight.

Lucario has Azen and Atomisk (who also uses a slew of other characters, including MK). Pikachu has Anther. Incidentally, Anther is also responsible for many of Sonic's "wins," and I guarantee you that for most of those, Pikachu played a greater role. The only other guy who won with him, DJ Browny, only went all Sonic in one tourney that had 14 entrants and a $9 entry fee.

That is why your results are meaningless. They fail, utterly and completely to take into account the reality of tournaments. When you take into account that a grand total of four or five of the best players in the country are responsible for this wins, often with the help of some of the more powerful characters in the game, your list utterly fails to represent anything meaningful.

YES. I forgot that secondaries could also influence this. And that the same people are going to win regardless of who they use as long as they are significantly better than everyone else.

Take a look at what your topic is trying to say and see if you can convince yourself that this actually make sense. If a character is good they will have many high placings. If a character is one of the best they will have many wins. The way that tournaments play out causes better characters to have better placings. There isn't some invisible wall that stops a character from getting many top 8s, but allows a character to get a few wins if they are played well enough. So, both in theory and practice, there are no characters that are better at getting wins than they are top 8 placements.

I don't think you have a useful topic.

Enjoy your worthless list.

 

[hyperstation]

This is so useless.

I'd love to know when and how you decided that the ratio of wins to overall top8 placements is equivalent to high-reward? That's rhetorical, in case you didn't know, so please...please don't answer me.

 

[Crizthakidd]

or we stick to anoukos list that gives a greatter data collection. why would u need this list that has snake and mk that low on their % to win when cleary they have won more tournies than any other char