Meta Tier List v2.0 Discussion & Moving Forward

[Gunla]

Hey all,

It's been a little over a week since v2.0 launched. No patch (thank god) and average groups made it a much smoother process than 1.0 in terms of releasing. Next week (or sometime around then) I am planning releasing a follow up and would like to hear some opinions of yours about the tier list (which would also likely be going into the article). Did you like the more loose format? Were you not a fan of it? How about some of the final placements? Reflections compared to v1.0?

Here's a list of the data (take note of standard deviation):

Spoiler: Raw Data

Character | Average Score | Standard Deviation (Score) | Average Group | Standard Deviation (Group)
Diddy Kong | 1.32 | 1.501 | 9.88 | 0.391
Cloud | 2.18 | 2.18 | 9.84 | 0.418
Sheik | 2.66 | 2.734 | 9.65 | 0.7
Rosalina | 3.39 | 2.085 | 9.66 | 0.634
Mario | 4.88 | 2.443 | 9.33 | 0.786
Sonic | 5.39 | 2.756 | 9.29 | 0.733
Fox | 5.88 | 2.376 | 9.15 | 0.776
Zero Suit Samus | 7.09 | 2.182 | 8.83 | 0.782
Ryu | 8.41 | 4.165 | 8.68 | 0.874
Mewtwo | 8.46 | 3.07 | 8.72 | 0.73
Bayonetta | 8.93 | 3.914 | 8.74 | 0.885
Pikachu | 13.88 | 3.274 | 7.67 | 0.9
Meta Knight | 14.54 | 5.931 | 7.62 | 1.092
Villager | 15.07 | 4.358 | 7.51 | 0.806
Mega Man | 15.93 | 5.837 | 7.4 | 0.937
Ness | 16.39 | 4.916 | 7.41 | 1.068
Toon Link | 16.64 | 3.739 | 7.23 | 0.836
Corrin | 17.85 | 6.337 | 7.15 | 1.107
Marth | 19.36 | 6.852 | 7.01 | 1.063
Captain Falcon | 20.95 | 4.615 | 6.72 | 0.922
Greninja | 21.11 | 5.731 | 6.8 | 1.052
Lucario | 21.52 | 5.405 | 6.63 | 0.968
Yoshi | 23.64 | 5.201 | 6.42 | 0.984
Pit | 23.73 | 6.264 | 6.39 | 1.115
Donkey Kong | 24.41 | 4.854 | 6.27 | 0.98
Peach | 24.52 | 5.898 | 6.2 | 1.079
Dark Pit | 24.89 | 5.888 | 6.27 | 1.044
R.O.B. | 25.64 | 5.393 | 6.08 | 1.17
Luigi | 27.34 | 6.583 | 5.95 | 1.126
Olimar | 30.93 | 7.914 | 5.62 | 1.338
Robin | 30.98 | 6.396 | 5.42 | 0.931
Bowser | 31.55 | 5.812 | 5.23 | 1.018
Lucas | 32.18 | 7.071 | 5.17 | 1.149
Wario | 32.61 | 4.471 | 5.26 | 0.943
Ike | 32.68 | 5.711 | 5.06 | 1.03
Lucina | 34.52 | 7.746 | 4.95 | 1.178
Mr. Game & Watch | 37.8 | 6.637 | 4.38 | 1.337
Pac-Man | 38.73 | 6.016 | 4.34 | 1.097
Little Mac | 38.8 | 6.891 | 4.21 | 1.149
Palutena | 38.91 | 5.881 | 4.11 | 1.156
Kirby | 39.27 | 6.413 | 4.23 | 1.166
Duck Hunt Dog | 40.0 | 7.697 | 4.21 | 1.303
Link | 40.64 | 8.186 | 4.09 | 1.338
Shulk | 41.77 | 6.753 | 3.86 | 1.267
Samus | 42.66 | 6.459 | 3.71 | 1.186
Wii Fit Trainer | 43.18 | 5.465 | 3.65 | 1.043
Bowser Jr. | 43.18 | 5.907 | 3.61 | 1.138
Dr. Mario | 45.16 | 4.554 | 3.47 | 1.051
Falco | 45.75 | 4.141 | 3.25 | 1.036
Roy | 46.02 | 5.323 | 3.38 | 1.191
Charizard | 48.27 | 3.763 | 2.82 | 1.077
King Dedede | 49.02 | 4.06 | 2.7 | 1.146
Mii Gunner | 51.26 | 6.322 | 2.28 | 1.124
Mii Brawler | 51.89 | 6.696 | 2.26 | 1.316
Zelda | 52.07 | 2.235 | 1.92 | 0.616
Ganondorf | 52.66 | 2.254 | 1.9 | 0.651
Mii Swordsman | 53.85 | 3.951 | 2.04 | 0.947
Jigglypuff | 53.98 | 1.433 | 1.72 | 0.604

Some of the biggest public concerns for characters included:
(Standard Deviation's Top 6)


- Placement compared to Marth



(Other notables)
- Overall treatment and usage of an addendum to help determine their place in the list
- Placement, why he wasn't next to Pit (regional opinions)


Moving forward, I think that doing a biannual tier list is a good idea (meaning we'd start voting in December and end probably around sometime at the end of January (maybe accommodating for Genesis 4 like we did last time). There's other things to also think about in-between that time (Matchup Chart potentiality, etc) as well.

 

[Luco]

Just gonna pitch in real quick and say Lucas should probably be *significantly* higher, and I personally take fault for not editing my vote in time to take into account Tai's recent success in the US along with generally rising results for the character (#nobias).

Now that that's out of the way, the process was in some ways really nice, but perhaps a standard calculation of all of our lists put together causes too much craziness. I have a suspicion it might be healthy from here on in to assume our tiers are roughly correct for now and discuss individual tiers - how characters could move within those tiers, if a character should be moved in or out of that tier and how many spots. Unless the meta has crazy changes within the next 6 months, this could allow us to work out more minor differences such as, 'Should Cloud be #2 or #8 (or anywhere within those numbers)?' or 'does DK now deserve to be placed in high tier versus his current position in mid tier?'

This would probably mean defining our tier groupings a little more, a process which I'm unfamiliar with. Perhaps any members of previous BRs could share how they defined their tier groupings for their lists?

Perhaps people feel our meta is changing too rapidly for this way of refining our future tier lists to be relevant right now, I dunno.

 

[Sethlon]

As far as I remember, we didn't do anything special for organization between tiers, for the BBR; we simply tallied up points for where every character was on all the lists that were put in and organized them that way, and then put cuts in the tiers based off of where there were big differences.

IMO, if anything, our list is TOO split...while the current list put out has a nice even number of characters in most sections and looks pretty, the cuts in most lists that were put in by 4BR members had fewer splits, which is more indicative of Smash4's rather compact balance.

 

[Fox Is Openly Deceptive]

IMO, if anything, our list is TOO split...while the current list put out has a nice even number of characters in most sections and looks pretty, the cuts in most lists that were put in by 4BR members had fewer splits, which is more indicative of Smash4's rather compact balance.

Agreed.
I also wasn't a fan of how it made it seem like there was a set number of characters that were allowed to be in a tier, especially considering that so many votes were deliberately more shaped like a bell curve.

There is one other thing I would like to note, perhaps next time we could work out which characters have the larger deviations before releasing the list to the public and perhaps we might then want to think about why such differences in opinion exist. If for example it becomes apparent that a character like Link is consistently voted higher by the representatives that know the character inside and out at a high level, one would hope that this would be taken note of and people would consider adjusting their votes accordingly if they weren't sure where that character should go.

 

[Catana]

I also didnt like how split the list was, and how it went all the way down to H tier (this isnt brawl). And why were there only 7 chars (14 combined) max in each tier? Like the game is definitely way more balanced than this, which has been shown several times, and it doesnt seem to match the majority of the votes made.

On top of that I also disagree with how the splits inbetween tiers were made, although this is just my personal opinion, but also regarding how balanced this game is I dont think you can just make a split each 2 tiers for the sake of making splits. What it looks like to me is the layout of just a generic tier list, instead of one specifically for this game. For example imo it should be more like;
S = top. A, B = high. C, D, E = mid. F = low.
I think we should've been given a preview of the tier list before it was made public (although Shaya made a post one might call a ''preview'' even though the chars werent shown, and it didnt seem to be very open for discussion despite it having been posted in the discussion thread), maybe voting should only be step 1 towards making a tier list instead of having that define the end result. After that we could discuss about it so we could refine it a little and make sure it makes a little more sense in case people here dont agree with it still.

 

[Shaya]

most of the voters had 7 groups,
and we ended up with 9 tiers total.

Beyond compacting S/A tier (something we considered) and how we divvy'd up B/C, it's just how things worked out [there isn't much data justification for anything below C to be brought closer together]. It wasn't for aesthetics to have 7 "max" character tiers.

Now a bit of the ever lasting confusion comes from what a tier list is meant to be and how other smash games presented them.
If there has to be a low tier, then what do we define it as? I personally see -Low- as the lower end of the group data rather than explicitly 'low tier', which people tend to correlate to 'trash' and 'unviable'.

S-X, as has been said multiple times, is not meant to be an indicator of power level in contrast to anything else (particularly other games/tier lists or this pigeonholing school-based paradigm of F = lowest tier), it's just the terms we've used for it. As much as it isn't perfect, it's the standard way we've represented tiers through BRs for a long time, so an alternative arbitrary but non-contextually unambiguous form is ... difficult and no definitely-better solution has been presented.
Going by how we split up the higher regions, S/S- and A+/A ("why have alternative negative/plus though!?!?!") could've been alternatives to better express things, but like... ... eh.

I'd like to be more forthcoming with the data in the future. We have some 'anal' precedents in relation to showing the full results before we're ready to release because of leaking incidents. But dumping the data without characters should still give people here an opportunity to less-subjectively discuss how they would split things up (which would've been taken into consideration by me had people done so), I wish I had done so earlier.

 

[Luco]

The rhetoric going around that F tier = trash and unviable actually bugs me. People have become disillusioned with how we choose to represent a spread of data. I wish there was a way we could counter this rhetoric so we don't have to list out tiers A+ / A / A- for the sake of "representing the game's balance".

I know some people really like that way of ordering tiers and they're welcome to, just something I feel clutters up a list more than it helps.

 

[Shaya]

Other trains of thought: 65.5% of the cast (excl miis) was on averaged placed in the top half of the voters tiered groups.

Bell (normal) curve, restricted to 7 groups:

--ONE-- 95%
7:
6+:
--TWO-- 68%
5+:
4+:
3:
2:
--THREE-- 95%
1+: +

I suppose the notable difference, assuming you'd take -one- as top, -three- as bottom, is that you'd probably be inclined to make '5+' high (same as now except we have it as two tiers), '2+' low, and '4+' / '3' (the notable classification difference) as middle.

In a bit of hindsight 20/20, my view of balance primarily in contrast to past iterations, have me think G&W, Mac and Palu are 'middle tier' characters, and I'm sure quite a few people dealing with duck hunt, kirby, etc would think strongly about them too.

5 equal splits of the group distribution (or roughly 1/3rd of the cast [high/top]/middle/[low/bot] each) wasn't necessarily ideal, but there are larger differences (and deviations) seen as we move from G&W's group downwards than 5/4. Although it seems most are taken up by the whole 'amount of tiers' thing more so than the divisions, for the sake of low tier tournaments G&W's tier being bumped up to middle (lower mid?) is arguable.

 

[Fox Is Openly Deceptive]

I'd like to be more forthcoming with the data in the future. We have some 'anal' precedents in relation to showing the full results before we're ready to release because of leaking incidents.

See I have to take issue with you there unfortunately. What exactly are we safeguarding by preventing leaks of a 'still subject to change' tier list? And is it really more valuable then making the official tier list as good as it might otherwise be? Should we really be avoiding the opportunity for discussions that might help improve the final product in the name of stopping leaks? What's the worst case scenario here? Maybe there's a post on reddit or something claiming to know what the tier list will be. Big deal.

 

[erico9001]

Definitely a shorter amount of time is good for creating a tier list, and I am in favor of December due to Winter Break, but the discussion should be ongoing 'oh did you see X character's placement at X?'

I noticed a lot of disapproval of the tier list this last time, and people seemed quick to dismiss it. I recognize the biases people can have, but I don't think we should ignore them either.

First, take notice of the variability we have been having in results compared to tier list placements. We have had several F and G placements into top 64 and top 32 in recent majors. We had an E, Dath's Robin, come in 3rd at Shine! LeeT's Jigglypuff, the character at the way bottom of the tier list, came in 33rd.

With the amount of variability/lack of consistency with characters' placements, having divisions between D and E tier seems almost misleading, if not arbitrary. How were the amount of sub-divisions and divisions decided on? Tiers should represent a definitive separation in competitive viability. But in our tier list, what does saying a character is in F in lieu of G represent? It's a level of precision that Smash 4's meta has not reached yet, and won't reach for a while if ever.

Smash 4 is not nearly as polarizing of a game as the previous installments, but we have more tiers. This should not be the case. The reason why we have such variability in placings might be that Smash 4 is just much more of a game of reads and mental play, not character-character dominance. As it stands, our tier list paints a much different picture of our game. The tier list should be getting wider, to represent how much more level the metagame is in terms of balance, not deeper. More characters should not mean more tier lists. The tier list would overall be improved in accuracy, with the added benefit of people liking it.

Can we justify this many divisions or tiers with this more level game?

 

[Diosdi]

I have to aggree with Erico. Smash 4 did a good job at balancing, so most charaters are "viable". The gap of "character viabiliity" (how good a character is) between Smash's S tieer and H tier is maybe the difference of viability between only Melee's B and C tier, with double the cast. I think we have to stop thinking on "if" a character is viable and start thinking "how" and "why" is it viable. We have seen top placements with lower tiers, like Tai in doubles, so i think it is very clear that there is character diversity iin viability.

Smash 4 is a different meta than Melee and Brawl. May disaggree with the TL because of so many jumps there were, but I see it as in like the meta has evolved so much in the past few months that Marth and Mega Man are high tiers. The high deviation only shows how diverse opinions are, and should not be taken as a set back.

I am in favor of bianual TLs. Voting int Dec allows us to see placements in this fall-winter and a mid summer votation allows us to see most majors like CEO, EVO, APEX, Genesis, etc.

In the matters of the topics that have alerted people I cant respond as I dont have the context of the discussion made for the TL.

 

[erico9001]

I have to aggree with Erico. Smash 4 did a good job at balancing, so most charaters are "viable". The gap of "character viabiliity" (how good a character is) between Smash's S tieer and H tier is maybe the difference of viability between only Melee's B and C tier, with double the cast. I think we have to stop thinking on "if" a character is viable and start thinking "how" and "why" is it viable. We have seen top placements with lower tiers, like Tai in doubles, so i think it is very clear that there is character diversity iin viability.

Smash 4 is a different meta than Melee and Brawl. May disaggree with the TL because of so many jumps there were, but I see it as in like the meta has evolved so much in the past few months that Marth and Mega Man are high tiers. The high deviation only shows how diverse opinions are, and should not be taken as a set back.

I am in favor of bianual TLs. Voting int Dec allows us to see placements in this fall-winter and a mid summer votation allows us to see most majors like CEO, EVO, APEX, Genesis, etc.

In the matters of the topics that have alerted people I cant respond as I dont have the context of the discussion made for the TL.

I was looking more into it, and there was a fair amount of discussion about this in the c2 tier list discussion thread, especially starting on the final page. This is the explanation:

Most of our voters provided lists with 7-8 tiers/groups, we have 9 tiers here.
Data could go either way for whether we have a split between S and A tier (Group score vs Weighted Group). We could have a bigger B tier ('high' is split by weighted score) and bump down 3 characters in C to 'top' of middle (reflecting group score).

The Top/High/Mid/Low/Bot were devised by the range of group scores (9.88 - 1.72) split equally into five. You would notice the tiers are mostly a reflection of 10-9, 9-8, etc of group score, however we bumped 3 (20th to 22nd) up as they were the only outliers produced from the 5 way split. We merged the bottom tier as well.

--------------------TOP--------------------
S:
A:
-------------------HIGH--------------------
B:
C:
--------------------MID---------------------
D:
E:
--------------------LOW-------------------
F:
G:
------------------BOTTOM----------------
H: ** *

If I am gathering correctly, the primary reason is most people had a fair amount of divisions in the tier lists, though even then the tier list has more than the average. Problem is I don't think people are really paying much attention to the amount of divisions they are putting in, just rather using the previous posts for the standard of how many divisions they will construct the list out of.
Anyways, we should probably have a serious discussion about it.

 

[Player-1]

I feel like the names of the tiers should be more of an indicator of how well the chars can perform in the meta as opposed to being relative. For example, if bottom tier is our lowest tier then Jigglypuff is bottom tier relative to the rest of the cast, but, at least for me, I don't see Jigglypuff as a "bottom tier" character, I see her more of a low tier based on how well she can perform in a tournament. The only character I would describe as a "bottom tier" char in the game is 1111 guest mii brawler.

 

[Diosdi]

IMO things like "the names" of the tiers are something not that relevant. Someone has to be the best character, and someone has to be the worst. That is just how it is.

If we have to sugar coat it for other to not feel insulted, fine, but in the end there is a best character, and there is a worst character - no matter how tiny the differences are.

For example, if we make a tier list between the Mario Bros (Mario, Luigi, and Doc) there are only three of them. One HAS to be the best, other HAS to be the midle and the other one HAS to be the best one. Applies the same relative rule to a cast of 58 chars: someon has to be on the top, and someone at the bottom

 

[Illusion.]

Some of these "inaccurate" placings for characters can be a result of our members not knowing exactly what the characters in question are capable of (MU spreads/theory) and the results they obtain, which is absolutely understandable since there are over 56 characters in the game and we all have lives outside of Smash.

I think it'd be a good idea if a week or two prior to voting, we go over each character and take a look at all the results they've gotten and their MU spreads/theory. For example, for , all the 4BR mains would get together to create a MU chart and compile all the results they've gotten using Das Koopa's data thread for all the voters to see.

This "database" would look something like this:

Spoiler: Mario

MU spread: *insert image here*
Results: *insert all results here from the current period*

Spoiler: Luigi

MU spread: *insert image here*
Results: *insert all results from current period here*

 

[Diosdi]

Some of these "inaccurate" placings for characters can be a result of our members not knowing exactly what the characters in question are capable of (MU spreads/theory) and the results they obtain, which is absolutely understandable since there are over 56 characters in the game and we all have lives outside of Smash.

I think it'd be a good idea if a week or two prior to voting, we go over each character and take a look at all the results they've gotten and their MU spreads/theory. For example, for , all the 4BR mains would get together to create a MU chart and compile all the results they've gotten using Das Koopa's data thread for all the voters to see.

This "database" would look something like this:

Spoiler: Mario

MU spread: *insert image here*
Results: *insert all results here from the current period*

Spoiler: Luigi

MU spread: *insert image here*
Results: *insert all results from current period here*

I think that for this to be done for all characters (so that all have a fair shot) would take too long. Doind on a day we would have to start like now for the next voting, and there are still tournaments between today and the next voting, si although it is a good idea, i think it would take too long. What i do think is that we can compile data from character discords, that have weekly JMUs, and although the JMUS are no way near finished, it would be a boost to our data collecting Following what Illusion said, ther should be a foreplay before voting. Pointing out the strenghs of each chara two or three weeks before voting, but not the MU spread which imo would take too long

 

[Illusion.]

It is a lot of work, but that's kinda the point of being in here; we volunteered to work.

Stuff like MU spreads can be done whenever (even now to get it out of the way) as long as it's done before voting, and results can be an ongoing process since new ones come in. Results would be easy to do since we have a reliable resource available to us.

 

[Diosdi]

Although, I would say we started to do an official MU spread; maybe even a voting on it. Like voting for the TL, people could just grab a format in a spreadsheet, fill, send and data could be averaged. I would start the discusion on a thread but i dunno how things work in here yet XD

 

[shrooby]

Anything involving MUs across the whole/majority of the cast wouldn't be a project done "here." It would involve communicating with character communities, like what was done for the Brawl MU chart. Because you can't rely on one or two people to give accurate rating descriptions of every MU. The only character I can think of that we have a decent number of representatives for is Greninja, go figure. (And there's plenty of characters for which we have zero.)
An MU chart would be a very large and very time consuming project, even if it was just half of the cast. It would not (and should not) be viewed as a side-project to the tier list.

Not to say that when we're voting you shouldn't try to educate us if you think a character is being under or overrated in votes. But any "official" MU-thingy would most likely not be limited to 4BR members.

 

[Sinister Slush]

Most likely not the case, but a MU chart has prolly been withheld cause patches and known DLC characters have been a thing for the past 2 years almost.
Last couple characters was released in February and the possible final patch hit 3 months after that in May. We're 3 months now since that patch so there could've been no way to finish a MU chart for every character.
Especially since Melee only had to deal with 26 characters, brawl 37, than Smash 4 over here with a whopping 58.

It might've been because smashboards was slowly dying out 3 years ago, but the final MU chart being worked on was incredibly slow with a lot of discussion never finishing or even being discussed at all. I think ROB Snake and Marth having the most members I believe but no posts being made to theorycaft and agree on a ratio anywhere, especially for the snake mains.
So who knows how it'd be to get everyone to get together again to do all this, especially with from what I read, there was a bit of a hassle to get tier list votes from a small chunk of 4BR people.

The process of doing a MU chart would also take about as long as doing a tier list, with maybe 3+ months more due to 21 characters added from 37 to 58.
Of course realistically not like Melee/Brawl BR discussed all 625/1296 matchups, but we still discussed a good portion over the 8+ years for the previous games.


I know where you're going with the 4br members of one character getting together to make a single character MU chart, but it'd still need to be discussed by other 4br members and even choose outside sources from panel leaders again just like the BBR mu chart days to discuss.
More opinions, more accuracy. Especially if 4br members are choosing the people, it shouldn't be hard to deal with someone that's stubborn.

 

[Diosdi]

I aggree with this. A MU chart should not be limited to the 4BR and as i said could be discussed with the char discords. This would probably not mean a patch, but on Sep 30th regular tournaments are being removed...although that could just be a maintanance and no patch. Im just warning.

Either way, the meta seems pretty stable rn, although it is very young, so if we start rn, for the time we finish there would have been discoveries that can change a big part of the work, so i say we wait a bit for doing a MU spread

 

[Shaya]

See I have to take issue with you there unfortunately. What exactly are we safeguarding by preventing leaks of a 'still subject to change' tier list? And is it really more valuable then making the official tier list as good as it might otherwise be? Should we really be avoiding the opportunity for discussions that might help improve the final product in the name of stopping leaks? What's the worst case scenario here? Maybe there's a post on reddit or something claiming to know what the tier list will be. Big deal.

If you mean having an 'in-progress' view of things that would increase the work load a lot (most people aren't providing things we can easily automate through and we'd have to pay attention to if/when people edit; this time around I was traveling overseas so I wasn't able to personally keep up with it), otherwise voting would generally be closed when revealed. Data was revealed without the final orders proper before the final closing though.
Also I am a fan of discussion.

I think that the anonymous/unnamed data; perhaps with high standard variation characters revealed, would be a nice venture that avoids the leaking issue and a bit of the subjective 'they must be X' that comes with character bias.
Also another notion that I brought up with Marc is tier list voting periods being... weeks, not months. If a hyper majority of voters are going to do it at the last minute / need to be reminded anyway, any longer period is, again, a larger workload and superfluous.

 

[---]

In regards to the MU chart, I'd say I'm more in favor of getting work started on that sooner rather than waiting for discoveries/patches. A MU chart would probably fill the niche the BR usually plays within the community better than a tier list at the current time, as well as potentially serving as a catalyst for future projects and the unspoken goals of increased transparency, attracting more people to SWF, anti-meta shenanigans, etc. It's a lot harder for some top player who wishes to promote themselves to disavow a full MU chart than it is for a single character or a tier list. Plus the BR has been kind of dead the past few months too...

The biggest hurdle I would think is actually getting it started with each character board involved. Unless we get some sort of patch or a NX announcement by end of the month/mid-October(?) we should start looking into this as our next project.

Also another notion that I brought up with Marc is tier list voting periods being... weeks, not months. If a hyper majority of voters are going to do it at the last minute / need to be reminded anyway, any longer period is, again, a larger workload and superfluous.

How long does it take to get votes from international players?

 

[Diosdi]

In regards to the MU chart, I'd say I'm more in favor of getting work started on that sooner rather than waiting for discoveries/patches. A MU chart would probably fill the niche the BR usually plays within the community better than a tier list at the current time, as well as potentially serving as a catalyst for future projects and the unspoken goals of increased transparency, attracting more people to SWF, anti-meta shenanigans, etc. It's a lot harder for some top player who wishes to promote themselves to disavow a full MU chart than it is for a single character or a tier list. Plus the BR has been kind of dead the past few months too...

The biggest hurdle I would think is actually getting it started with each character board involved. Unless we get some sort of patch or a NX announcement by end of the month/mid-October(?) we should start looking into this as our next project.



How long does it take to get votes from international players?

I think that it should be the other way around. Working and then receive a patch or omething that changes relevant stuff would probably kill the work we had done before the patch. and an MU chart is an extensive work that needs a long timespan; in which patches or game-changing discoveries might come up. An MU chart needs dozens of studies, all of the ddifferent, and having something that may alter factors can result in wasted time. So maybe, we should wait a bit for an MU chart

 

[Gunla]

On that note, reading Sakurai's latest column, the new maintenance that didn't come with a patch, and the last patch looking like a hot fix makes me confident that patches probably won't be a concern for the future, unless a game-breaking bug is found.

How long does it take to get votes from international players?

Usually a case-by-case scenario. Japan votes come from Vayseth and usually come as a group (last time around I got them all on the deadline day, just a matter of reminding). European votes seemed to come much faster (iStudying, while he was waiting to post till the last day, did work on it for an extended period of time).

Looking into it, I do think that shortening the voting periods should likely be done; people either get on the ball right away or wait till the deadline. Middle period is very light on votes and wastes time.

As for the MU chart business, my personal proposal for the moment is to maybe start with a "Top Tiers" MU chart, and gradually work from there based on internal feedback.

 

[Player-1]

What was wrong with the the way the BBR did the MU chart project? It seemed to work out well.

 

[Illusion.]

By a MU chart, I didn't mean an official one that includes every character like what you can find here. I meant individual MU charts, so the Greninjas would submit one for their character, the Yoshis would submit one for their character, and etc.

But as someone mentioned earlier, not every character is represented in the 4BR so getting character Discords and sub-forums to help would be essential.

 

[A10theHero]

Conveniently enough, I actually started surveying Pika mains in the Pikachu Discord about 's MUs recently. People often visit there asking if there's a generally-accepted Pikachu MU chart, and we don't really have something to offer them. So that's why I'm doing this. c:
I've been using Google Forms for that. It's really easy to work with, so I encourage everyone to try it. You have the ability to export the data directly into a spreadsheet in Google Sheets, so that's good too. Something I definitely recommend is telling them to use the +/- # system when putting in their opinions on MU ratios. It's significantly easier to work with in a spreadsheet. (I learned that the hard way. >.<) One of my friends there told me you can modify the spreadsheet to automatically factor in new responses and update MU ratios. I can't do that for my chart since people writing it like 50:50 or something would cause errors. :/ But you can learn from my mistakes!
Once you have the data, you have the choice of taking either the average or the median. The median is less susceptible to the effects of outliers/unpopular opinions and it is more likely to be a number like +1, -2, etc. The average can show more nuance in the responses than the median (for example, instead of the result being +1, it could be something like +.6594, meaning that opinions are more in the middle of even and slightly advantageous, rather than being solidly in one category). Both have their respective pros and cons, so it actually could be better to calculate both and compare them.
Something I've kinda been on the fence about is factoring in the MU charts of top players. If I were to do that, I'd probably be contacting ESAM, Captain L, Rideae, Tachyon, Z, Kenny, and Pika4Life. ESAM of course has interesting opinions and he's not the only one (that goes for all characters, really). Though, with a larger pool of data, extreme opinions wouldn't be as impactful, so maybe if I get a certain number of responses, I'll add in their opinions. Determining that "certain number" is the current problem for me. That number would probably be different for every character since they have different amounts of mains and top players. Another thing I could try is just making two different charts--one that factors in top player opinions and one that doesn't. Of course, a separate chart for the average and the median would mean that there would be 4 different charts.
It's also worth mentioning the limitations of this method. Using mine as an example, because I've only been asking Pikas in the Discord group to survey so far, the results technically cannot be generalized to the actual population of Pikachu mains. Also, because I'm not doing a random sample and I'm just letting people voluntarily fill out the form, I'm introducing a lot of bias to the results. Specifically asking top players for their opinions is actually kinda problematic too. All things considered though, we really don't have the resources to collect data in a way that perfectly follows statistical principles. It's still worth thinking about, but we can't really do more than that. We can only do the best we can. ¯\_(ツ)_/¯

(I don't think I emphasized this enough, but this is just what I happen to be doing. That doesn't mean this is exactly what we should do or we have to do this. It's all up for discussion. I'm just trying to contribute what I can. Hopefully that's clear enough. ^-^)

 

[#HBC | Red Ryu]

What was wrong with the the way the BBR did the MU chart project? It seemed to work out well.

Only real issue that occurred was the lack of discussion from some character mains and zero knowledge of how "tie breakers" where handled in some situations.

 

[CT Chia]

The MU chart project was such a headache in trying to resolve conflicting opinions which usually just resulted in going to which side was more stubborn.

 

[A10theHero]

The MU chart project was such a headache in trying to resolve conflicting opinions which usually just resulted in going to which side was more stubborn.

Couldn't that be avoided by simply surveying character mains and collecting data on their opinions? All we'd do is just combine the data from both sides and then take the average and/or median.

 

[CT Chia]

We did try to include knowledgeable community members of said characters that weren't in the BR. Activity was an all time low.

 

[Player-1]

I think that varied from character to character. I was in the Diddy group of the BBR MU chart and I know we were a fairly active group compared others. Some of the characters didn't have a lot of representation at higher levels of play. Take Kirby for example, the only relevant kirby was pretty much chudat (with mike kirby being the next best I think, but there was quite a significant skill gap). So I don't recall Kirby's MU discussions getting a lot of activity there.

I also think that a lot of the top level players that were in the MU chart discussion felt like that they didn't want to waste their time discussing MUs with players that were significantly worse than them and relied on theorycrafting for MUs (not that I agreed with them, it's just what I felt like why a lot of top level players didn't want to be active). I think that's what happened to the MK group.

But I think overall the MU chart ended up well and got a lot of good publicity aside from the 1st one and also created a lot of discussion which was also good.

 

[Gunla]

Hey all,

Discussion seems to be generally wrapping up and we want to release the followup in the next week. For this followup, we'll be covering the following characters in more detail:

• (highest standard deviation)
  
• (2nd highest standard deviation)
  
• (Position relative to Marth)
  
• (General position)
  
• (Position relative to Pit)
  
• (General Position)
  
• (27 position jump)

If any of you would like to lend your voice for the followup article, please contact me.

(Pre-emptively tagging Luco and Myran for Lucas and Olimar)

 

[Luco]

Gunla Did you want me to compose a small write-up for Lucas? I'm happy to do that, although I want to make sure I'm speaking for more than just myself: To everyone else active here, I want to ask why you placed Lucas where you put him and whether that would mirror your opinion now? If not, why didn't you edit your vote to include your change of opinion?

Keep in mind "time / laziness" are perfectly legitimate responses. "I still think Lucas is middling" is also very fine (although try to give some context on this one). The important thing is to be honest so I know I'm speaking for all of us.

I might also go PM a few people on twitter who aren't so active here.

 

[Myran]

So Olimar is a character that most players including some Olimar players consider bad. His results aren't as strong as other characters, but I think that's due to other circumstances as opposed to the characters flaws. Most top Olimar's including myself don't travel out of their states that often. For various reasons, mine being money, this means he doesn't get very strong displays often. He's not perfect, but he's top 15 imo, and easily high tier. I do generally well in FL almost always landing top 8 at the big tournaments here, and have been for awhile. I know Shuton does very well in Japan even making top 3 at some of the stacked ones. There's some other good placings, but I don't know them all.

What I think really holds Olimar back in many players eyes is that they either have interactions with him that skew their perspective and make them thing oh he's not good. (Looking at the regions with Olimar players who may not be as strong) Information regarding him that has circulated, but is only a part of his overall character. The information I'm talking about is how Pikmin clank. Yes they are considered projectiles. However, they have some interesting properties that allow them to circumvent the issue of clanking when performed right. To add to this they're some of the best projectiles in the game only really losing to ones that can't be stopped. Examples being Thoron, fully charged Water Shuriken, and ROB Laser.

So let me break down some of the techniques and things Olimar can do to circumvent his largest issue, the clanking. There's something that I coined a "desync smash". This happens when Olimar is moving and starts a smash after the movement. So if he's running and does an up-smash or a pivot fsmash, or if he's falling and lands with a smash (Usually dsmash) it can trigger it. What happens is if the opponent throws out and attack it can catch the Pikmin during the start of the smash before it's hitbox comes out, it's hurtbox is out because it has been desynced by the smash. By the opponent throwing out an attack it locks the Pikmins position in place (This means the Pikmin will smash from where it was as opposed to where Olimar is currently), and finishes the smash. (Once a smash is started it can't be stopped or killed during the attack) Since the opponents attack hit the Pikmin before the smash hitbox came out the opponent is stuck there in the animation. This allows the smash to commence uninterrupted hitting them. I've punished some of the best moves in the game with this, True Shoryuken being one of them. It does sound situational, but with continued practice I'm getting better and better at making it happen. Since it's really just awareness and positioning it's a viable tool to punish attacks you cannot normally. On a smaller not you can do this technique to a similar extent without desyncing the position. Whenever a smash is being charged the Pikmin sits their with it's hurtbox active, this means when spaced right it can soak hits then release hitting the opponent. It's how Olimar can punish things like Shuttle Loop, Mario Up-B, Shoryuken, and more under the ledge when they have hitboxes. Actually if the moves trade at the right time Olimar may even be hit, but the smash will still hit them since their move hit something. This is just one of the tools Olimar has to work around his weaknesses.

(Don't have a handy video for desync smashes, but if it's needed I can make one.)

Another big advantage he can gain is lagging his opponents hitbox's with his Pikmin. This can backfire at times, but I'd say a solid 90% of the time it's allowed me to punish moves I couldn't normally. It's also allowed me to dodge ones I shouldn't have been able too. This is an extremely important use of the Pikmin, and is again something I feel more Olimar players should put into practice.

He can pressure extremely well with the proper lineup. Having two to three purples can allow him to pressure shield better than most if not all characters in the game imo. Retreating purple bair is safe on shield, and when coupled with purple side-b before hand it can allow for strong pressure. His smashes are also notorious for poking shields. (I hear this complaint from players to often to ignore it) His dsmash can break shield when up close since both Pikmin can hit shield, and with a red and purple or two purples it's not hard to take it from 100% to 0%. His nair is a great way to snub out the approaching opponents move since it's a nice long multihit.

His defensive options also help him reduce the issue of being juggled somewhat easily. He has whistle which has super armor frames 6-12 and very little lag allowing you to tank a hit, and then punish most of the opponents moves. B-Reverse Pikmin Pluck allows him to switch his momentum and direction similar to what whistle did in Brawl. His up-b can be used as a third jump and immediately canceled into an aerial if need be. (It can also get out of combos and multihits often). He's hard to gimp since his recovery covers so much distance, and he can chuck purples towards you to put of some hitboxes while he recovers.

What I'm getting at is that he has strong tools to combat the opponent, and he can be the aggressor with the proper lineup. The biggest thing holding him back is how unforgiving he is to play. He requires more than just managing Pikmin, his spacings and timings have to be very precise most of the time. Unlike most characters, being off makes you just outright lose the transaction or have a Pikmin do a standard desync. (Most desyncs can be worked around with proper spacing and awareness) It's a work in progress from me and other top Olimars. I think once we start attending more it'll make him look better to most, but for the time people are seeing what is a work in progress. He doesn't have amazing confirms like ZSS, Ryu, or others. He has some, and they're strong, just not as over-centralizing as theirs. I'm one to look at Olimar's abilities when people can actually put his tools into proper use. I've had days where I felt I could take any character in the game no problems just by using his tools properly. So if people wanna tell me why they feel he's mid tier or worse I'd love to hear it. I don't think saying results really does it justice, because in the states where he's present he usually does well. He just doesn't get to attend nationals that often.

 

[Gunla]

Luco
Something probably a bit similar to the first followup. You could speak on an individual level on why you think he ended up where he did, or maybe how you felt on the general voting trends, as well as how you think it might change in the future.

Or do something akin to what Myran has done - something of a decent length is perfectly acceptable.

 

[M@v]

Here's my writeup. Gunla feel free to edit and tinker with as needed. Its a work in progress and I'll either be finishing it tonight or tomorrow.

Lucina's placement compared to Marth is an issue that is brought up quite often, and it is again the case with the new tier list. Why is Lucina, a character that is almost identical to Marth, not right next to him/tied with him on the tier list? Before we dive into this with more detail, let's start off by pointing out in earlier patches, Marth and Lucina were indeed much closer to eachother. But overtime, the list of buffs they received helped Marth more in the longrun than Lucina. This is in addition to the inherent advantages the tipper mechanic has in many situations. Let's break it down:

Let me preface by briefly covering how their swords work.
Lucina does not have a tipper, but her sword does what is approximately inbetween a marth tipper and non tipper in terms of damage and knockback along the entire blade (I can hunt the exact stats later; pretty sure they are in the Marcina boards). This brings up the fact that Lucina gets more damage out any attack that is not tippered compared to Marth. For a good while, this was the cornerstone of the pro Lucina argument: Unless you were perfect, Lucina could potentially be more beneficial in the longrun since you will not be landing every single attack as a tipper, no matter how good you are as a player.

With that aside, lets dive in.

First, the patch changes that affected Marth and Lucina. There were multiple patch changes that affected them roughly the same (IE slight % increases in damage here and there), but there were two patch changes in particular that helped Marth especially.

The big one: The shield stun changes.
When the patch rolled around that introduced increased shield stun (I need to go back and double check which one), this had a massive impact on Marth. Tippers induce a lot more shield stun on an opponent. The result of this is that Marth can be much safer than Lucina when attacking a shielding opponent. With proper spacing, his tippers moves can be much harder for a shielding opponent to punish. Having this huge safety advantage over Lucina pays massive dividends, especially when having to face down top tier characters where a couple frames can mean the difference between being safe and being punished.

The 2nd change is the downthrow change. This change allows Marth and Lucina to use down throw as a more effect combo starter and 50-50 creator. The difference is Marth gained true tipper combos from this, giving him another guaranteed way to deal out maximized damage. This is something Lucina doesn't gain since she has the same hitbox all around.

All of Marth and Lucina's A moves follow the same principle for the most part: Marth's will be better if spaced, Lucina's will be superior if not. Marth can gain some better combos though by hitting with untippered moves, which will lead into tippered follow ups easier.

Their grabs and throws are exactly identical. Both do the same things at the same %s

Their Special moves is where it has been argued Lucina has an advantage. But does she really? Depends.

(B) Shield breaker: Although Lucina's is still a powerful move, Marth's give the strongest chance to break shield while charging the least amount of time due to tippers. Both have stock ending potential though.

(Side B) Dancing blade: In general, Lucina's is much more reliable in terms of KO potential. The last hit of Lucina's side b is pretty strong and can end stocks at decent percents. That being said, Marth's can tipper more often than people realize. Following an opponents DI and ending it with a side direction or up direction attack pending their DI can net you a tipper quite often, as well as just bad DI from an opponent. As you can imagine, this can net super early kills.

(up B) Dolphin slash This move is pretty much better for lucina in most situations. Most times this move is used to recover, a combo finisher, or as an out of shield option where it is very hard to control spacing. Lucinas will result in more consistent KOs.

(down b) counter: Again, a toss up. Marth's will have the edge here though I believe, as in a lot of situations you should be able to not only read an opponent's move to counter, but properly space yourself to hit with a tipper. Sometimes though, spacing is impossible.

Lucina also has a fatal flaw in her design. In most matchups, despite having the ability to do consistent damage anywhere on the blade, she has to play the spacing game anyway to be the safest. So essentially, so has to play like marth and go for tippers to remain as safe as possible anyway. In these situations, you are being directly gimped by not being Marth in a matchup where you have to play like Marth. A straight downgrade.

Last but not least, results. This section speaks for itself. Marth has much better and consistent results in the current meta. Players my Mr.E and Pugwest have put up solid results with Marth, top 16ing major events. Meanwhile, Lucina is nowhere to be seen in those higher placings.

 

[Sethlon]

IMO things like "the names" of the tiers are something not that relevant. Someone has to be the best character, and someone has to be the worst. That is just how it is.

I couldn't disagree more. While there obviously has to be a highest tier and a lowest tier (by definition of a tier list, the labels of each tier describe a lot in what sort of difference is between each tier. Using a completely different letter for each tier, as we did for the current list, implies a good amount of gap between each tier, which I think a lot of us agree isn't the case; Smash4 has perhaps the best balance of any Smash game to date, and its tiers should be more condensed accordingly. Brawl was a horridly imbalanced game, as far as the entire cast was concerned, and it only went out to 'F' tier. Ours should be labelled to show the difference.

Personally, I like the idea of using this list;

--ONE-- 95%
7:
6+:
--TWO-- 68%
5+:
4+:
3:
2:
--THREE-- 95%
1+: +

- but rather than numbering them or having a letter for each tier itself, using partial letter jumps with either '+'s or '-'s to show the difference isn't that huge. So something like;

--ONE-- 95%
S+:
S:
--TWO-- 68%
A+:
A:
B+:
B:
--THREE-- 95%
C: +

Or perhaps using 'S+, S A+, A, B, C, D' to show how the curve dips as far as the lowest characters are placed. I imagine you get the idea

 

[Gunla]

Thank you to those who responded and wrote for the article. I really do appreciate it.

You can read it here: https://smashboards.com/threads/4br-tier-list-v2-0-followup.441975/