Official Competitive Character Impressions 2.0

[SKX31]

Results for Smash Ultimate Summit 3.

1. Tweek
2. MkLeo
3. Sparg0
4. Marss
5. Zackray
5. VoiD
7. Kola
7. Aaron
9. Charliedaking
9. Maister
9. Dabuz
9. Riddles
13. Atomsk
13. Cosmos
13. KEN
13. Dark Wizzy

Tweek, Leo and Sparg0 were playing out of their minds this weekend.

A thing I noticed with Sparg0 in particular is that his sets vs. top 10 caliber opponents have been hit-and-miss:

SWT CA Regional:

• Mkleo (Winners Semis) 2 - 3: Game 1 was close - 120 % - but Leo got a jump-in read to seal the game. While Sparg0 took Games 2 and 4 comfortably, he also lost a major percentage lead mid-way through Game 3 - Leo was at 2 stock 146 % to Sparg0's 2 stock 47 % - but then Sparg0 got kill-hungry / nervous and Leo turned the game on its head. When Sparg0 got that kill - at 188 % - Leo had already taken Sparg0's second stock and tacked on 60 %. Sparg0 made it back but lost Game 3 off of a really low DI'd back-throw. Game 5 was Leo leading from the start and not relinquishing it.
• Maister (Losers Quarters) 2 - 3 (reverse swept): This one was rough for Sparg0. He took a close Game 1 and convincingly snagged Game 2. Game 3 was really rough though - Sparg0 had a 2-to-1 stock lead, 0 % on both players, when Maister clawed his way back and sealed the game off of Sparg0 barely missing a Pyra F-Tilt, got grabbed and then chased to the ledge... where Sparg0 was caught by Maister's Dash Attack and lost the game. Sparg0 fell behind early in Game 4 but clawed back... only to see his potentially set-sealing D-Air ledgetrap barely whiff and Maister capitalize on a whiffed Prominence Revolt seconds later with a D-Smash. Game 5 was mostly even until last stock, where Maister ran roughshot over Sparg0.

Summit:

• Tweek (Group stage) 3 - 1. Only asterisk here is that this came on Tweek's off day (when he also lost to Aaron 0-3), but an important win since it meant Sparg0 didn't run the risk of facing Leo in Winner's Quarters. Game 2 was a two-stock affair, which helped.
• Zackray (Winners semis) 3 -0: While this has two small asterisks (the Zackray SD at Game 1's beginning and him panic-switching to Wario in Game 3) it nevertheless was arguably the cleanest set from Sparg0. Beating Zackray in the same place he lost to Leo a week prior is still a good sign.
• Tweek (Winners finals) 1 - 3: Had a solid lead in Game 2, with two stocks and Tweek at around 100 %. Got too kill-hungry / nervous, and that allowed Tweek to take neutral interaction after neutral interaction. Did have a similar lead in Game 3 but sealed that to cut Tweek's lead in the set to 1 - 2. Yes, Sparg0 can pride himself on being the only one to take a game off of Tweek in the bracket - but there's another near-miss in Game 4 where Sparg0 barely missed an Up Smash conversion that would've at least scared Tweek and made that game a very close call... and got punished for missing that by Tweek.
• Mkleo (Losers finals) 2 - 3 (reverse swept): Sparg0 clawed back from a tough situation in both Games 1 and 2 to win both games... and in a kinda cruel twist of fate was close to making a similar comeback in Game 3, down to 1 stock vs. 3 Leo stocks. Only Leo's second stock just wouldn't go down - Leo tacked on 100 % ending in a DBZ trade before he got KOed, and Leo took Game 3 as a result. Game 4 was kinda close until Leo took Sparg0's second stock and made Sparg0 go for desperate plays. Game 5 was even, with it being sealed by the exact kind of conversion that seem to elude Sparg0 in those decisive moments: an Arsene Up Smash confirm.

I really think his primary issue right now is the nervousness - that combined with his tendency to swing caused him to slowly but surely lose leads and thusly several games. It's why I chose to highlight the sets he lost - it's a consistent theme. On the upside he has the time to work on that and he has the potential to become a Top 10 player - it's just that he needs some more work to get there IMHO.

 

[Hydreigonfan01]

Initial seeding for Riptide
https://www.reddit.com/r/smashbros/comments/pfa8ax

 

[Thinkaman]

When doing big picture review of OrionRank numbers, I like looking at Log2(n+1) of the character scores. This measure might seem to overly marginalize the differences among the top while magnifying wild swings at the bottom, but if you're not using it for week-to-week comparisons I find it to be a very good filter holistically.

Here is what it looks like normalized out of 10: (Normalizing it doesn't change anything else)

(As a reminder, be careful comparing across OrionRanks too much--they are real-world samples that are neither equal-sized nor controlled.)

Viewed logarithmically, OrionRank Phase 1 actually meets the formal criteria for a normal distribution! This speaks to the quality of this filter choice, and of course Das Koopa 's methodology + data.

Phase 2 and 2021 are almost a normal distribution, but have become slightly too near-top-heavy; there are too many scores clustered in the high tier section. Playing around with the spreadsheet, both would indeed qualify as a normal distribution if any 2 of a couple dozen "high tier" characters were deleted.

OrionRank is great because its deeper-cut weighting collects more data and diffuses more of the results down into the high and mid tiers than more superficial measures. I would postulate that an even deeper cut measure than OrionRank--beyond the scope of our available data collection, methodology, and Das Koopa's sanity--would diffuse results lower still, to the point that the results would again be log-normally distributed. This is backed up by any number of broader usage trends we have looked at, like the smash.gg dataset or either SWT analysis. (All of which correlate with OrionRank 2020 and 2021)

Pure logarithmic view; Mac, Ganon, and Swordfighter are currently off the bottom of this scale.

Phased differently, I think that Ultimate's weighted results appear to be log-normally distributed in general, only faintly confounded by an organic/unavoidable numerical bias in OrionRank to high-tiers in the "sweet spot" of its data collection range. There's not really an actionable takeaway from that, I just found it interesting.


The biggest news here honestly is, while people talk about Aegis and Diddy, continues dominating. According to OrionRank, his current deviation in results is greater than any previous #1 character at any point in Ultimate's life. For comparison, his #2 is performing roughly as far above everyone else as Joker was at his peak. (Pre-nerfs Joker!)

What's more, they aren't just popular. Palutena, Snake, Ness, Bowser, and (historically) Cloud are the characters we think of when we picture characters whose performance is as much represented by the legion of players who choose to play them as their peak results. ROB and Roy both dramatically outperform their usage levels, which in ROB's case are already top 5. In this lens, Roy almost looks like "the new Wario."

 

[Aaron1997]

The biggest news here honestly is, while people talk about Aegis and Diddy, continues dominating. According to OrionRank, his current deviation in results is greater than any previous #1 character at any point in Ultimate's life. For comparison, his #2 is performing roughly as far above everyone else as Joker was at his peak. (Pre-nerfs Joker!)

What's more, they aren't just popular. Palutena, Snake, Ness, Bowser, and (historically) Cloud are the characters we think of when we picture characters whose performance is as much represented by the legion of players who choose to play them as their peak results. ROB and Roy both dramatically outperform their usage levels, which in ROB's case are already top 5. In this lens, Roy almost looks like "the new Wario."

I find it insane that statistically ROB is the best character in the game but Tier list and Super Major results don't reflect even close to this. It really shows that the character choices the absolute tippy top player make is more important to deciding the best then statistics.

 

[Nemesis561]

A thing I noticed with Sparg0 in particular is that his sets vs. top 10 caliber opponents have been hit-and-miss:

SWT CA Regional:

• Mkleo (Winners Semis) 2 - 3: Game 1 was close - 120 % - but Leo got a jump-in read to seal the game. While Sparg0 took Games 2 and 4 comfortably, he also lost a major percentage lead mid-way through Game 3 - Leo was at 2 stock 146 % to Sparg0's 2 stock 47 % - but then Sparg0 got kill-hungry / nervous and Leo turned the game on its head. When Sparg0 got that kill - at 188 % - Leo had already taken Sparg0's second stock and tacked on 60 %. Sparg0 made it back but lost Game 3 off of a really low DI'd back-throw. Game 5 was Leo leading from the start and not relinquishing it.
• Maister (Losers Quarters) 2 - 3 (reverse swept): This one was rough for Sparg0. He took a close Game 1 and convincingly snagged Game 2. Game 3 was really rough though - Sparg0 had a 2-to-1 stock lead, 0 % on both players, when Maister clawed his way back and sealed the game off of Sparg0 barely missing a Pyra F-Tilt, got grabbed and then chased to the ledge... where Sparg0 was caught by Maister's Dash Attack and lost the game. Sparg0 fell behind early in Game 4 but clawed back... only to see his potentially set-sealing D-Air ledgetrap barely whiff and Maister capitalize on a whiffed Prominence Revolt seconds later with a D-Smash. Game 5 was mostly even until last stock, where Maister ran roughshot over Sparg0.

Summit:

• Tweek (Group stage) 3 - 1. Only asterisk here is that this came on Tweek's off day (when he also lost to Aaron 0-3), but an important win since it meant Sparg0 didn't run the risk of facing Leo in Winner's Quarters. Game 2 was a two-stock affair, which helped.
• Zackray (Winners semis) 3 -0: While this has two small asterisks (the Zackray SD at Game 1's beginning and him panic-switching to Wario in Game 3) it nevertheless was arguably the cleanest set from Sparg0. Beating Zackray in the same place he lost to Leo a week prior is still a good sign.
• Tweek (Winners finals) 1 - 3: Had a solid lead in Game 2, with two stocks and Tweek at around 100 %. Got too kill-hungry / nervous, and that allowed Tweek to take neutral interaction after neutral interaction. Did have a similar lead in Game 3 but sealed that to cut Tweek's lead in the set to 1 - 2. Yes, Sparg0 can pride himself on being the only one to take a game off of Tweek in the bracket - but there's another near-miss in Game 4 where Sparg0 barely missed an Up Smash conversion that would've at least scared Tweek and made that game a very close call... and got punished for missing that by Tweek.
• Mkleo (Losers finals) 2 - 3 (reverse swept): Sparg0 clawed back from a tough situation in both Games 1 and 2 to win both games... and in a kinda cruel twist of fate was close to making a similar comeback in Game 3, down to 1 stock vs. 3 Leo stocks. Only Leo's second stock just wouldn't go down - Leo tacked on 100 % ending in a DBZ trade before he got KOed, and Leo took Game 3 as a result. Game 4 was kinda close until Leo took Sparg0's second stock and made Sparg0 go for desperate plays. Game 5 was even, with it being sealed by the exact kind of conversion that seem to elude Sparg0 in those decisive moments: an Arsene Up Smash confirm.

I really think his primary issue right now is the nervousness - that combined with his tendency to swing caused him to slowly but surely lose leads and thusly several games. It's why I chose to highlight the sets he lost - it's a consistent theme. On the upside he has the time to work on that and he has the potential to become a Top 10 player - it's just that he needs some more work to get there IMHO.

Its pretty clear that Sparg0 is a top 10 player right now, most believed he was in that realm of skill level coming into the post quarantine era and he only helped those beliefs with his performance at these 2 events. Sure he has to continue to place well to cement this but just off the eye test and listening to the other top players who have played him, it's very clear

 

[Hydreigonfan01]

Its pretty clear that Sparg0 is a top 10 player right now, most believed he was in that realm of skill level coming into the post quarantine era and he only helped those beliefs with his performance at these 2 events. Sure he has to continue to place well to cement this but just off the eye test and listening to the other top players who have played him, it's very clear

Yeah, I think you've also got to look at the positives of his play at Summit too. He was the only player at the event who put Leo to a game 5 situation, the only one to take Tweek to game 4 in the final bracket and also beat several consecutive top 10-top 15 players (Aside from Zackray and Tweek, KEN is a top 15 player and Riddles might be top 20). He also beat Maister at a smaller event a while ago. There's a reason why Leo and Tweek have so much respect for him, and they're considered top 2 in the world.

 

[SKX31]

Its pretty clear that Sparg0 is a top 10 player right now, most believed he was in that realm of skill level coming into the post quarantine era and he only helped those beliefs with his performance at these 2 events. Sure he has to continue to place well to cement this but just off the eye test and listening to the other top players who have played him, it's very clear

Yeah, I think you've also got to look at the positives of his play at Summit too. He was the only player at the event who put Leo to a game 5 situation, the only one to take Tweek to game 4 in the final bracket and also beat several consecutive top 10-top 15 players. He also beat Maister at a smaller event a while ago. There's a reason why Leo and Tweek have so much respect for him, and they're considered top 2 in the world.

I can certainly see the argument for him being top 10 - I should clarify. I'm just personally someone who views Top 10 as a very, very high bar requiring longer-term performance - and we've only seen two major offline tournaments with him / six major sets vs. people considered Top 10 since the restart. That and the difficulty closing out sets has me a bit wary.

Still though, I'm not trying to deny that he's an extremely good player - he is, not many can compete with the absolute best with a lot of close results in tow. And I'm reiterating that I can see the argument because of that. A lot of those losses were also by very slight margins, and that does signal that he has progressed a lot in a year, and could possibly reach (or really confirm that he is within) that area should he continue to perform like this.

When doing big picture review of OrionRank numbers, I like looking at Log2(n+1) of the character scores. This measure might seem to overly marginalize the differences among the top while magnifying wild swings at the bottom, but if you're not using it for week-to-week comparisons I find it to be a very good filter holistically.

Here is what it looks like normalized out of 10: (Normalizing it doesn't change anything else)

View attachment 328722

(As a reminder, be careful comparing across OrionRanks too much--they are real-world samples that are neither equal-sized nor controlled.)

Viewed logarithmically, OrionRank Phase 1 actually meets the formal criteria for a normal distribution! This speaks to the quality of this filter choice, and of course Das Koopa 's methodology + data.

Phase 2 and 2021 are almost a normal distribution, but have become slightly too near-top-heavy; there are too many scores clustered in the high tier section. Playing around with the spreadsheet, both would indeed qualify as a normal distribution if any 2 of a couple dozen "high tier" characters were deleted.

OrionRank is great because its deeper-cut weighting collects more data and diffuses more of the results down into the high and mid tiers than more superficial measures. I would postulate that an even deeper cut measure than OrionRank--beyond the scope of our available data collection, methodology, and Das Koopa's sanity--would diffuse results lower still, to the point that the results would again be log-normally distributed. This is backed up by any number of broader usage trends we have looked at, like the smash.gg dataset or either SWT analysis. (All of which correlate with OrionRank 2020 and 2021)

View attachment 328725

Pure logarithmic view; Mac, Ganon, and Swordfighter are currently off the bottom of this scale.

Phased differently, I think that Ultimate's weighted results appear to be log-normally distributed in general, only faintly confounded by an organic/unavoidable numerical bias in OrionRank to high-tiers in the "sweet spot" of its data collection range. There's not really an actionable takeaway from that, I just found it interesting.


The biggest news here honestly is, while people talk about Aegis and Diddy, continues dominating. According to OrionRank, his current deviation in results is greater than any previous #1 character at any point in Ultimate's life. For comparison, his #2 is performing roughly as far above everyone else as Joker was at his peak. (Pre-nerfs Joker!)

What's more, they aren't just popular. Palutena, Snake, Ness, Bowser, and (historically) Cloud are the characters we think of when we picture characters whose performance is as much represented by the legion of players who choose to play them as their peak results. ROB and Roy both dramatically outperform their usage levels, which in ROB's case are already top 5. In this lens, Roy almost looks like "the new Wario."

Huh, now this is really interesting. I have a question regarding the characters straddling the y-line: If lets say Scend, LeoN, Elegant or Larry Lurr start to take more high placings at (super)majors, how much may it impact the OrionStats? Or would it take multiple mains of one of those characters to start significantly affecting the character's placement? For instance saw a major jump between Orion 1 and 3, but I'm not entirely sure that's mainly due to Maister or him + other mains contributing together.

I'm asking partly since multiple line-straddlers (in particular , , , and ) have similar tier placements according to this doc that averages tiers*, and I'd imagine that any of those characters getting a shift upward in OrionStats could result in an improved tier list perception generally.

*(With the caveats that some characters std. deviation are larger than others, particularily Marth, and that some tier lists have huge singular tiers, like Tweek's)

 

[Thinkaman]

It's just that you can measure character performance at different levels of play. Characters are going to look different at top 1mil players vs top 10,000 vs top 100 vs top 1.

The problem is that attempting to measure performance at a top 1, top 10, or even really top 100 level says a lot more about individual players than the game itself or broader social phenomenon in the community. It's just too volatile and tiny-sample to mean much of anything. OrionRank dips down deeper into top 1000 territory, which is still extremely high but has both a lot firmer ground to stand on and a lot more applicability to other levels of play.

It is far more likely that your locals see more ROBs than Diddys.

 

[Aaron1997]

There was a problem fetching the tweet

New Platform Kazuya tech

 

[Hydreigonfan01]

PGStats did a video on the big DLC characters that people should be worried about at Riptide, and and .

 

[Lacrimosa]

Temple just started and the first upset did already happen as well:

Noxumbra 0-2 LeyKoO ( ) (209th seed) (Not sure who they usually play, last online tournament was with Shulk but they also played Joker at one point in the past). Given their Twitter profile it's more likely it was Shulk.

E: Twitch chat says Shulk.

Wouldn't bother with that usually, but it's the same Richter that achieved 4th place last WANTED which was fairly big.

Other (notable) pool upsets:

Pool A (Winners):
BeJay 0-2 Kace
Pew ( ) 0-2 Yakumo09
Greil ( )1-2 Nibodax
Homika 0-2 AndresFN /

Pool A (Losers):

BeJay 0 - 2 Pew (out at 97th) (That part of the bracket really got upsetted a lot since seed 33 played 32 for 97th place)
Pew 0-2 LeMonke (out at 65th)
Noxumbra 0-2 otek (out at 65th)


Pool B (Winners):

Sintro 1-2 Hazhell
LuiS 1-2 Demo
Mr. E 0-2 Naskino
TriM 1-2 Rinor

Pool B (Losers)

Sintro 0-2 Momon (out at 129th as the 29th seed)
LuiS 1-2 Nishinoma (out at 129th as the 24th seed)


Pool C (Winners):

Ramses 0-2 MoDzai
Longo 1-2 Myollnir (?)

No further big upsets in losers in Pool C.


Pool D (Winners):
Supahsemmie 1-2 Otakuni
DarkThunder 0-2 Mezcaul

 

[Hydreigonfan01]

Not entirely related to character impressions but I do think this needs to be brought up.

Temple is apparently crowded as ****, there's way too many people for the Venue to allow, it stinks and people are looking to DQ including Mr. R.

There was a problem fetching the tweet

Looks like the OrionStats ranking for Temple is going to go lower due to the poor quality of the tournament.

There was a problem fetching the tweet

This is just a COVID epidemic waiting to happen.

 

[Nobie]

So I’ve had some time lately to lab Mewtwo, and I’ve found something interesting with how the Teleport start-up interacts with ledges.

Basically, when Mewtwo leans up out of a dash or does that spinny stop out of a run, it actually alters the starting position of Teleport. For example, normally, if you Teleport back while standing at the ledge, it travels the length you expect. With the shift, however, you Teleport as if in the air.

If you look at the Teleport effects at the beginning, they're at slightly different spots. Also, there’s no need to run off or get airborne, meaning that if you choose NOT to Teleport, you're still grounded.

I took that idea of that slightly shifted Teleport, and found that if you combine it with the c-stick-shortened Teleport tech, you get some interesting things, like this on-stage Teleport ledge trump:

Again, the fact that you can do it out of a grounded position instead of running off to get the trump is potentially useful. The fact that you don’t have to be facing away from the blast zone is a plus as well.

Using the shortened and shifted Teleport also allows for an easier ledge cancel dair from Battlefield platforms, at least in terms of angling (you only have to press straight down on the c-stick to get this reliably):

I don't think it's gonna be a game changer or anything, but it can be another tool in Mewtwo’s already large kit, and it might make execution easier in certain situations.

 

[Hydreigonfan01]

Dabuz did a Rosalina & Luma MU chart

There was a problem fetching the tweet

 

[WatwatBreton]

mewtwo stuff

Yo this is sick! How easy/consistent are they to do? I'd love to have some practical teleportation applications coz that move is so cool but "you have to be within these pixels otherwise you dead" current cancels are a bit whack lol.

 

[Nobie]

Yo this is sick! How easy/consistent are they to do? I'd love to have some practical teleportation applications coz that move is so cool but "you have to be within these pixels otherwise you dead" current cancels are a bit whack lol.

As someone who stinks at doing precise Teleport angles and is also afraid of sliding offstage and dying, I find that the angles are quite a bit easier in these cases. The hard part is more a matter of timing, especially in consistently executing shortened Teleports when that's necessary.

As a reminder, short teleports/short warping:

 

[Thinkaman]

Dabuz did a Rosalina & Luma MU chart

There was a problem fetching the tweet

Broke: "My character is distinctly above-average according to matchups but should be buffed anyway."

Woke: "My character was played less than Random in the SWT qualifiers and should be buffed."

 

[RonNewcomb]

Not entirely related to character impressions but I do think this needs to be brought up.

Temple is apparently crowded as ****, there's way too many people for the Venue to allow, it stinks and people are looking to DQ including Mr. R.

There was a problem fetching the tweet

Looks like the OrionStats ranking for Temple is going to go lower due to the poor quality of the tournament.

There was a problem fetching the tweet

This is just a COVID epidemic waiting to happen.

I've always been impressed by how this doesn't happen more often. Just thinking about weighing how much money I as a TO would have to put out on choosing a venue that's not too big to break the bank & make the event feel empty but not too small to prevent this kind of disaster stresses me out to no end.

And of course no one ever, ever pre-registers which would've helped a ton with planning.

I feel bad for both the players and the TO and the poor staff.

 

[Thinkaman]

Just for fun, temple top 64 character count:





Missing:




...and all DLC besides Joker:

Note: I over-simplified all players to a single main. I think there are a couple players with meaningful secondaries, including a couple using otherwise missing characters like Terry. Also some of this might be outdated--for all I know half the lower bracket players I've never heard of have switched mains to Sephiroth and Mii Gunner.

 

[ParanoidDrone]

Broke: "My character is distinctly above-average according to matchups but should be buffed anyway."

Woke: "My character was played less than Random in the SWT qualifiers and should be buffed."

So I'm amazingly OOTL as regards the details of Rosalina's meta -- what exactly changed about her to make her viability plummet in ultimate? Like, I can tell that something's different just from playing her, but apart from the obvious like her new nair I'm casual enough these days that I can't quite put my finger on what that something is.

 

[sleepy_Nex]

Thinkaman Lucky mains Pythra now. He also played her onstream if i'm not mistaken. Idk wether he used pit or not so far.

 

[Thinkaman]

So I'm amazingly OOTL as regards the details of Rosalina's meta -- what exactly changed about her to make her viability plummet in ultimate? Like, I can tell that something's different just from playing her, but apart from the obvious like her new nair I'm casual enough these days that I can't quite put my finger on what that something is.

I know little about Rosa, but from my vantage point the big thing was Uair juggles. Smash 4 Rosa uair was one of the best moves in the series, a huge hitbox lasting 22 frames (27 with Luma!) she could spam every 42. The late hit basically set up a repeat strong hit for free, on a characters whose movement was setup to exploit this to the maximum extent possible.

Enough of that. In Ultimate, it's still huge, but only lasts 12 frames (14 with Luma) on a move that takes 50. The late hit has a lot more knockback growth, so it doesn't self-combo as much. Furthermore, this is in a game with both faster movement and directional air dodges. Suddenly Rosa's golden ticket is, well, just a good uair.

It's wild how buffed this character is otherwise. Higher weight. Faster Luma respawn. An entirely new nair that is faster, auto-cancels, and hits in front of her first. Fair is faster and SHACs too. Faster/safer specials. Better tilts, safer dash attack, jab is frame 5 instead of 8.

But none of this offsets how crazy Smash 4 uair was, and how much it single-handedly carried this character. I just pulled up a random Dabuz Rosa Smash 4 match and started counting uairs--it was over half of his attempted attacks in the stock I watched. (And far more than half of the attacks that landed.)

 

[Swamp Sensei]

So I'm amazingly OOTL as regards the details of Rosalina's meta -- what exactly changed about her to make her viability plummet in ultimate? Like, I can tell that something's different just from playing her, but apart from the obvious like her new nair I'm casual enough these days that I can't quite put my finger on what that something is.

I think the general game speed getting faster made it harder for Rosalina to do the specific spacing the character thrives in.

I kind of view them like Marth, on paper, they're still great characters, but the Ultimate engine screwed them over.

 

[The_Bookworm]

So I'm amazingly OOTL as regards the details of Rosalina's meta -- what exactly changed about her to make her viability plummet in ultimate? Like, I can tell that something's different just from playing her, but apart from the obvious like her new nair I'm casual enough these days that I can't quite put my finger on what that something is.

I know little about Rosa, but from my vantage point the big thing was Uair juggles. Smash 4 Rosa uair was one of the best moves in the series, a huge hitbox lasting 22 frames (27 with Luma!) she could spam every 42. The late hit basically set up a repeat strong hit for free, on a characters whose movement was setup to exploit this to the maximum extent possible.

Enough of that. In Ultimate, it's still huge, but only lasts 12 frames (14 with Luma) on a move that takes 50. The late hit has a lot more knockback growth, so it doesn't self-combo as much. Furthermore, this is in a game with both faster movement and directional air dodges. Suddenly Rosa's golden ticket is, well, just a good uair.

It's wild how buffed this character is otherwise. Higher weight. Faster Luma respawn. An entirely new nair that is faster, auto-cancels, and hits in front of her first. Fair is faster and SHACs too. Faster/safer specials. Better tilts, safer dash attack, jab is frame 5 instead of 8.

But none of this offsets how crazy Smash 4 uair was, and how much it single-handedly carried this character. I just pulled up a random Dabuz Rosa Smash 4 match and started counting uairs--it was over half of his attempted attacks in the stock I watched. (And far more than half of the attacks that landed.)

I think the general game speed getting faster made it harder for Rosalina to do the specific spacing the character thrives in.

I kind of view them like Marth, on paper, they're still great characters, but the Ultimate engine screwed them over.

The gist is that got some critical aspects of her gameplan from SSB4 nerfed into Ultimate, the gameplan that allowed her to be top 4 in SSB4.

Jab, a big part of SSB4 Rosalina, got significantly nerfed.
The big one is that the movement and range on Luma's jab has both been reduced. This greatly nerfs three important aspects about the character: 1) it hurts her ability to stop approaches, as Luma was a more active threat, and punishing Luma, especially with it taking lower knockback in SSB4, made it harder to shake it off; 2) it hurts her ability to zone the opponent out, for reasons similar to stopping approaches; 3) most importantly, it hurts her ledge trap game, as she was easily SSB4's best, most deadly ledge-trapper, as Rosalina could simply jab at the ledge at a safe distance, while opponent has to respect it, especially since Luma's jab itself was more deadly.
In the transition, jab 3 got its knockback significantly reduced. Especially with rage, Luma's jab 3 had surprisingly strong knockback. As with most rapid jabs in the transition from SSB4 to Ultimate, rapid jab got its damage nerfed quite a bit, which hurts this character in the particular due to how she works.
The rapid jab can now KO as well of the start-up of the jab being decreased (note that this only applies to Rosalina's jab, not Luma's), which are cool, but it does not compensate for everything else mentioned.

New neutral air is cool, but the old one with its bigger hitbox, higher active frames, and less startup on Luma also had contribute heavily for the duo's ability to completely wall opponents out.

Oh yeah since I am in the subject of it, Luma takes much less knockback in SSB4 than in Ultimate. Ultimate attempts to compensate this by making Luma's respawn timer slightly lower, while Rosalina's more general aggressive playstyle in the transition helps her fight out more independently. While it does somewhat help compensate, Luma taking more knockback in Ultimate, thus easier to dispose and ward off, hurts the character significantly.

Gravitational Pull also had more startup and less active frames. While it does have 5 frames less endlag, in SSB4, the active frames long exceeded the endlag of the move for whatever reason. Now the active frames ends 1 frame before the endlag of the move. The move also has less effective range. In general, it is harder to use for its purpose.

As mentioned above, Luma's up air juggling got worsened thanks to its higher startup and endlag, and lower duration, while the nerf to rage hurts it ability to just randomly KO the opponent. Similarly, Luma's down air got its startup and endlag significantly increased. While the duo's down air is still a solid landing option, combined with the changes to Ultimate's airdodges, the duo is more vulnerable to juggling, kinda similar to resident tall floaty character Mewtwo (who also coincidentally got a weight buff in the transition).

You kinda see the general pattern here: thrived in playing very defensively like puppet characters typically do. Their walling capabilities in SSB4 was frankly absurd thanks to the above mentioned tools. These walling capabilities allowed the duo, when played efficiently, to completely shut down many characters in SSB4, and play toe-to-toe with most top tier characters very efficiently. Even characters typically considered higher tiered, such as Luigi, DK, and Ness, gets completely walled and shut down by the duo. Cloud was Rosalina's main problematic matchup, but even then, I have seen Rosa players pull through. As such, it is no stranger that the duo is widely considered top 4 by the end of Ultimate's lifespan alongside Bayo, Cloud, and Sheik.

In comparison, essentially got reworked to play a far more aggressive playstyle. While they can play somewhat defensively, it is no where near as good as in the previous game, while their buffs more gear towards a more aggressive playstyle. Thanks to their higher weight and other buffed aspects of the duo, they can do so, especially after all the buffs in game updates, as the duo were very bad (borderline dysfunctional) characters at launch. Due to the more aggressive nature of Ultimate, I say that this was a smart move.
The issue is that this new more aggressive playstyle is much less safe than the old one, and kinda goes against their inherent design as a puppet character. Their new playstyle doesn't necessarily fix anything regarding their problematic matchups, as they are just as vulnerable, if not even more vulnerable, to the same stuff from the previous game. The duo frankly also has quite a bit of absurd stuff in the previous game that they lost in transition to Ultimate.

Hopefully that paints a picture on why the duo is noticeably less effective than in the previous game. I think they are pretty good, but no where near the heights of the previous game, where jab + neutral air alone can shut down so many character's ability to play the game.

 

[Lacrimosa]

There was a problem fetching the tweet

After the backlash at Temple, the TOs organized another hall to play in + additional set-ups.

That's how you fix a sub-par tournament experience. Good stuff.



Top 64 just started and quite some upsets already:

Round 1:
Oryon 1-3 Myollnir (?)
Tilde 1-3 AndresFN /
NickC 0-3 MoDzai
Flow 0-3 Otakuni
iStudying 2-3 Mezcaul





Losers:
DarkThunder 1-3 otek (out at 49th as the 15th seed)
Mr. E 1-3 Ramses (out at 49th as the 9th seed)


Some other early eliminations (not necessarily upsets because a lot of big players fell early):

NickC 3-0 Greil
Supahsemmie 3-0 Ogey
Homika 3-0 Dakpo (all out at 33rd)
Tilde 3-0 NickC
Raflow 3-0 Oryon (both out at 25th)

 

[Hydreigonfan01]

Fatality posted his thoughts on some characters

There was a problem fetching the tweet

 

[Lacrimosa]

Winners Top 8 qualifiers for Temple:

Kome 3-0 LeoN
KEN 3-0 Charliedaking
AndresFN 3-2 Glutonny (!)
Tea 3-0 Mr. R

Japan's too strong.


Loser Top 8 Qualifier:

Glutonny 3-0 Myollnir
LeoN 3-1 Supahsemmie
Raflow 3-1 Mr. R
Charliedaking 3-0 iStudying

 

[Nobie]

I found out Mewtwo can get a fairly reliable ledge cancel on Smashville by dropping down slightly from the platform before Teleporting. I'm not sure of the exact angle, but it's not hard to get, and this doesn't require any of shifting (i.e. kara canceling) or shortening I mentioned in that previous post.

The thing about Mewtwo is that it has a much harder time ledge canceling compared to Palutena in Ultimate, so I think this stuff has potential to make it a more reliable and approachable part of its game plan—especially for less precise players.

 

[Arthur97]

Not sure I buy Steve as top. Nothing really proving it yet, and there's been time to develop the blocky ones.

 

[Hydreigonfan01]

Not sure I buy Steve as top. Nothing really proving it yet, and there's been time to develop the blocky ones.

My guess is Fatality has personal bias, he is in the exact same region as DDee, who is looking to be top 4 in Georgia atm.

 

[Lacrimosa]

Top 8 for Temple looks a bit different than many would have expected:

Winners:
Kome - Tea
AndresFN - Ken


Losers:

Glutonny - Charliedaking
Raflow - LeoN



Out at 9th place, Mr. R seemingly wants to try out Pyra and Mythra more:

There was a problem fetching the tweet

 

[blackghost]

Fatality posted his thoughts on some characters

There was a problem fetching the tweet

most of this seems way off to me except the wolf and MAYBE aegis.
then again, never thought of georgia has having a diverse character pool. its the land of chroy.

 

[Ziodyne 21]

At this point Tea can very well rival Zackray and Protobaham as the #1 Player in Japan going by results..and a Top 10 player for sure.


may see a rise if rises in the meta a seems lime a near hard-counter to the latter

 

[Envoy of Chaos]

Speaking of Pac Man. Mew2King recently mentioned that he thinks Pac-Man could possibly be the best character in the game.

I initially rebuked the thought, but as I thought about it more... you know I started to think that idea might not be so absurd. When you look at Pac-Man’s kit and his strengths, I honesty failed to find a weakness anywhere on this character.

He’s got great normals, with good frame data that all complement his projectile game well, combo in themselves and let him box up close as well as any close ranged fighter. He’s got a fantastic shield and great OOS. He’s got damage output, he can kill outright with strong disjointed smash attacks he can honestly throw out at will and combo into it.

He has fantastic projectiles, they let him zone, control the pace of neutral. Literally lock a section of the stage off, give him a landing tool, kill, combo, setup into kills and plenty of trickery with sneaky bell or fruit setups combined with his hydrant or trampoline. Honestly I cannot think of another character with projectiles that enable them to do SO much with them.

Character can recover from just about anywhere off stage. Can go high, can go low, can stall, has multiple chances to get back to stage even if he’s being disrupted when getting back to stage. He can chase you anywhere off stage as a result. He can get off ledges and his ledge trapping is some of the best in the game.

I can’t think of a character in this game with such a complete kit. Joker, Pikachu, the Aegis. None of them are good in every aspect of the game the way Pac-Man is. I’m curious of everyone’s thoughts on Pac-Man, could he actually be the best character in the game?

 

[meleebrawler]

Speaking of Pac Man. Mew2King recently mentioned that he thinks Pac-Man could possibly be the best character in the game.

I initially rebuked the thought, but as I thought about it more... you know I started to think that idea might not be so absurd. When you look at Pac-Man’s kit and his strengths, I honesty failed to find a weakness anywhere on this character.

He’s got great normals, with good frame data that all complement his projectile game well, combo in themselves and let him box up close as well as any close ranged fighter. He’s got a fantastic shield and great OOS. He’s got damage output, he can kill outright with strong disjointed smash attacks he can honestly throw out at will and combo into it.

He has fantastic projectiles, they let him zone, control the pace of neutral. Literally lock a section of the stage off, give him a landing tool, kill, combo, setup into kills and plenty of trickery with sneaky bell or fruit setups combined with his hydrant or trampoline. Honestly I cannot think of another character with projectiles that enable them to do SO much with them.

Character can recover from just about anywhere off stage. Can go high, can go low, can stall, has multiple chances to get back to stage even if he’s being disrupted when getting back to stage. He can chase you anywhere off stage as a result. He can get off ledges and his ledge trapping is some of the best in the game.

I can’t think of a character in this game with such a complete kit. Joker, Pikachu, the Aegis. None of them are good in every aspect of the game the way Pac-Man is. I’m curious of everyone’s thoughts on Pac-Man, could he actually be the best character in the game?

His weakness is that half of his kit can be thrown back in his face without any specialized tools needed. Oh, and difficulty I guess.

 

[Rocketjay8]

His weakness is that half of his kit can be thrown back in his face without any specialized tools needed. Oh, and difficulty I guess.

I also recall him having marthritis problems after his combos stop working.

 

[The_Bookworm]

Temple: Hermes Edition

1st: Tea
2nd: Glutonny
3rd: KEN
4th: LeoN
5th: Kome
5th: AndresFn
7th: Charliedaking
7th: Raflow
9th: Mr. R
9th: Supahsemmie
9th: iStudying
9th: Myollnir
13th: Nitox
13th: Tilde
13th: Neeroz
13th: Mezcaul
17th: Naskino
17th: MoDzai
17th: Space
17th: Leon
17th: Flow
17th: Otakuni
17th: Tarik
17th: sisqui


According to SmashWiki, Tea used Kazuya at one point, although it is probably him messing around at pools.

Otherwise, this tournament did some big stuff on stabilizing some character's places in the meta:

• After the long silence since the online era started, finally gained some traction thanks to Glutonny's strong performance here.
• The Japanese trio of Tea, KEN, and Kome curb-stomped their way into top 8. Arguably the best user of each respective character, especially for Pac-Man and Shulk, these three made their presence known, with Tea winning the whole tournament.
• After mixed success online, LeoN continues to place very high in tournaments ever since offline returned. This combined with HERO's strong performances in Japan online, Bowser's strength as a character is undeniable.
• Thanks to AndresFn's performance in this tournament, taking some notable upsets, 's presence in the meta remains strong.

 

[Hydreigonfan01]

most of this seems way off to me except the wolf and MAYBE aegis.
then again, never thought of georgia has having a diverse character pool. its the land of chroy.

There’s only two Roy players in Georgia, Kola and Mugen.

 

[Aaron1997]

Speaking of Pac Man. Mew2King recently mentioned that he thinks Pac-Man could possibly be the best character in the game.

I initially rebuked the thought, but as I thought about it more... you know I started to think that idea might not be so absurd. When you look at Pac-Man’s kit and his strengths, I honesty failed to find a weakness anywhere on this character.

He’s got great normals, with good frame data that all complement his projectile game well, combo in themselves and let him box up close as well as any close ranged fighter. He’s got a fantastic shield and great OOS. He’s got damage output, he can kill outright with strong disjointed smash attacks he can honestly throw out at will and combo into it.

He has fantastic projectiles, they let him zone, control the pace of neutral. Literally lock a section of the stage off, give him a landing tool, kill, combo, setup into kills and plenty of trickery with sneaky bell or fruit setups combined with his hydrant or trampoline. Honestly I cannot think of another character with projectiles that enable them to do SO much with them.

Character can recover from just about anywhere off stage. Can go high, can go low, can stall, has multiple chances to get back to stage even if he’s being disrupted when getting back to stage. He can chase you anywhere off stage as a result. He can get off ledges and his ledge trapping is some of the best in the game.

I can’t think of a character in this game with such a complete kit. Joker, Pikachu, the Aegis. None of them are good in every aspect of the game the way Pac-Man is. I’m curious of everyone’s thoughts on Pac-Man, could he actually be the best character in the game?

Pac's weakness is that he's not (supposed to be) a threat without fruit. All his stuff takes time to get going and getting the fruit he needs in hand can be difficult without his opponent contesting him and threating to use them against him. Missing Bell confirm is such a huge swing in your opponents favor especially in timeout scenario's. It the few seconds it takes to get bell back into his hand is a huge opening for your opponent to get a edge. Without bell he can't kill you until 150+ unless he gets you with a raw smash.

I think Tea, the player is the reason why it looks like Pac has no weaknesses. He might just have the best advantage state in the world right now. He takes the smallest opening and runs a marathon with it, fruit or not. Its so good that he can camp 90% of the match but still have the illusion of being aggressive. Larry said yesterday he knows how to get the most out of Hydrant and always having at the right spot at the right time. He also doesn't suffer from bad habits that other Pac make like predictable disadvantage and tunneling on getting fruits.

Personally I think he's bottom of top 10. To many MU's that feel like a Up hill battle () and can even struggle vs some mid tiers like

TLDR: Tea is broken. Pac is as well but not as ridiculous as some people think.

 

[Hydreigonfan01]

Pac's weakness is that he's not (supposed to be) a threat without fruit. All his stuff takes time to get going and getting the fruit he needs in hand can be difficult without his opponent contesting him and threating to use them against him. Missing Bell confirm is such a huge swing in your opponents favor especially in timeout scenario's. It the few seconds it takes to get bell back into his hand is a huge opening for your opponent to get a edge. Without bell he can't kill you until 150+ unless he gets you with a raw smash.

I think Tea, the player is the reason why it looks like Pac has no weaknesses. He might just have the best advantage state in the world right now. He takes the smallest opening and runs a marathon with it, fruit or not. Its so good that he can camp 90% of the match but still have the illusion of being aggressive. Larry said yesterday he knows how to get the most out of Hydrant and always having at the right spot at the right time. He also doesn't suffer from bad habits that other Pac make like predictable disadvantage and tunneling on getting fruits.

Personally I think he's bottom of top 10. To many MU's that feel like a Up hill battle () and can even struggle vs some mid tiers like

TLDR: Tea is broken. Pac is as well but not as ridiculous as some people think.

Honestly Tea is looking to be a top 5 contender alongside Leo, Tweek, Zackray and Marss (Who knows about Proto, Glutonny or Sparg0 they could hit it too if they travel more and be consistent enough).
Edit: Anyway, Scend did a tier list

There was a problem fetching the tweet