ROB frame data discussion

[EGsmash]

TLDR: ROB's new Dthrow is terrible, and here's why

For some sad, salty reason I've been obsessed with proving to all who would listen that ROB's new down throw is actually a really terrible move, a huge nerf from Smash 4, and should never be used in serious play, ever. So I started labbing out frame data around the bury, how easy it is to mash out, First Active Frames for ROB and his Opponent, etc. While there is still the other thread on his Dthrow, I felt that this thread could be more about actual frame data (which seems an appropriate subject for our robot buddy).

METHODS:

• I used training mode to count frames
• All frame data is counted from the first bury frame, which is the first frame that the opponent receives damage from the throw
• Perfect Button Mashing is obtained by pressing one button one frame, and releasing it the next frame. Additional inputs do not achieve better results. Although it may feel more "natural" to alternate between two different inputs, all you're really doing is pressing and releasing one button on a 2-frame cycle. I happened to alternate Attack and Special for this lab because it was easier for me to count, but I did confirm that the press-and-release method works just as effectively. (Theoretically a turbo controller set at 30Hz would achieve this perfect 2-frame mash cycle, but that's probably not tournament legal)
• While I did lab out frame data for No Mashing and Perfect Button Mashing, I did not exactly figure out how much individual mash inputs affect bury time. It seems to be somewhat logarithmic, but I'm not sure.

ADDITIONAL DISCOVERIES:

• All bury moves in the game appear to have the same 'pop-up' lag frames before the opponent gets First Active Frame (FAF). This value is set at 12 with FAF at the 13th frame from the initial pop-up frame.
• CPU Shuffle does NOT affect bury or pop-up frames
• Stale moves do NOT affect bury or pop-up frames (aside from damage %)
• Rage does NOT affect bury or pop-up frames, nor does rage affect throw damage at all
• Character weight/size/fall speed does NOT affect bury or pop-up frames
• Button mashing before the initial bury frame does NOT affect bury frames
• Character weight DOES affect the height of pop-up
• Overall, Perfect Button Mashing reduces the duration of all bury moves by approximately 87.5%, or a reduction to 1/8th of the original duration.

FORMULAS:

PBM = Perfect Button Mash​

NM = No Mash​

FAF = First Active Frame​

D% = Damage% (Opponent)​

BF = Bury Frames​

Pop = Pop-up frames = 12​

Hit = Frame which a particular move's hitbox does damage​

​

ROB's Dthrow:​

​

NM:​

BF = (D% * 0.4) + 30​

​

PBM:​

BF = (D% * 0.05) + 5​

​

Frames counted from the first bury frame of Dthrow at which ROB can perform certain actions:​

​

FAF-ROB = 27​

ROB-Shield = 27​

ROB-SpotDodge = 30​

Hit-Dtilt = 29​

Hit-Jab = 29​

Hit-Utilt = 30 and 38​

Hit-USmash = 40​

​

Opponent Frame Data Counted From First Bury Frame

D%        NM        FAF-NM        PBM        FAF-PBM

0        30        43            5        18
25        40        53            6        19
50        50        63            7        20
75        60        73            8-9        21-22
100        70        83            10        23
125        80        93            11        24
150        90        103            12-13    25-26
175        100        113            14        27
200        110        123            15        28

FINDINGS:

1. Because opponents get FAF at 18 while ROB doesn't get FAF until 27, I was initially suspicious that, at 0%, Dthrow was truly punishable. However after testing characters with fast down attacks, ROB is always able to get his shield up on frame 27 at least a couple frames quicker than the opponent can actually hit him. This is due to the fact that even heavy characters pop up in the air enough so that the 9 frame distance gets eaten up by both start-up lag and travel distance. BUT...
2. This is a VERY short window for some characters (ZSS, Ganondorf, Dedede, etc). We're talking like 4-10 frames to get your shield up after FAF, which isn't really enough time to react to anything, especially considering input lag and the fact that your opponent can buffer moves out of pop-up better than ROB can buffer attacks out of Dthrow.
3. USmash is being considered ROB's early kill move out of Dthrow, but it doesn't come out until frame 40 AT MINIMUM. While this will hit an opponent at 25% who isn't so much as touching the controller, a Level 9 CPU could theoretically PBM as high as 300% or more, depending on character weight.
4. ROB's Utilt reaches fairly high and while the initial hit comes out at frame 30, the highest hit does not come out until frame 38. By the time opponents get FAF out of PBM, only the heaviest would be low enough to actually get hit by anything ROB can throw out quickly. Even at 200% damage where the opponent gets FAF at 28, they would likely be too high to be caught by the frame 30 hit, and characters like Ganondorf or Dedede have an advantageous trade with Utilt around frame 38.
5. ROB's Dthrow CANNOT 100% CONFIRM into ANYTHING until 225% when he can get you with Dtilt at frame 29. By then you should have just used Uthrow already

FURTHER RESEARCH and COMPARISON

Oh I didn't stop there. I also took a look at K. Rool's Dthrow. I won't post all the tables and junk, but I do have the formulas and 0% data:

K. Rool's Dthrow:​

NM:​

BF = (D% * 0.5) + 90​

​

PBM:​

BF = (D% * 0.0625) + 13​

​

FAF-KRool = 35​

​

Opponent Frame Data Counted From First Bury Frame

D%        NM        FAF-NM        PBM        FAF-PBM

0        90        103            13        26

​

1. K. Rool does have a longer bury overall at 0% with NM at 90 frames vs ROB's 30 frames.
2. At 0% with PBM, both ROB and K. Rool have about the same FAF disadvantage of 9 frames. However, this FAF spread evens out for K. Rool at around 140%, while ROB doesn't even out with his opponent's FAF until 175%
3. PBM against K Rool requires significantly more button inputs (12 versus 4 for ROB). This increases the likelihood of failing or buffering an unwanted move.
4. K Rool's bury moves come out quicker and do not telegraph nearly as much as ROB's, making it far easier to catch opponents off guard and delay effective PBM

TAKEAWAYS:

I know that human players are not going to be able to PBM perfectly every single time, so ROB's Dthrow can still be viable at higher % as a kill move as long as its used within a larger mixup game to fatigue opponents and catch them off guard. But given the huge telegraph when ROB does the move it makes it all the more easy for opponents to get their mash game on. Even at very high % attacking out of Dthrow is VERY UNSAFE and VERY PUNISHABLE if your opponent knows what they're doing. What's most saddening to me is that the hype around this move prerelease made it sound so amazing, but losing reliable, safe Beep-Boop combos out of Dthrow was just not worth what we got in return. Looking forward to future meta, I don't see a whole lot of use for the move aside from mind games and trying to possibly get your opponent to panic and maybe throw out something punishable. But that isn't really that fun imho.

I don't think bury moves are all that bad since they can all be PBM'd out of if you know what you're doing. However, ROB's bury move is REALLY, REALLY bad in light of that and deserves a buff to make it useful again. Either you could increase the Pop-up frames (which would be a global change), or reduce the effectiveness of PBM (perhaps doubling the initial 5 frame floor to 10 frames at 0%, which would increase the PBM button inputs required to 9) would put it more in line with how dangerous other characters' bury moves are.

But that's enough from me. Please someone else start talking so I can take a break!

 

[NGW]

Thanks for putting the time to put this together, good to see some numbers and data behind it. Is a shame, like I said in my other thread on his throw, I was split on how I felt about it, but as the weeks have gone by it really has underperformed to the point I only really use it once or twice a match, and even then only if I think I can seal a win out of opponents that may not be able to react to it (kinda like K. Rool's reflector, which seems best as a surprise option you use once and then have as a tool for mindgames). I did feel we may be losing more than we were gaining with this back when we first saw the change, but people tried to convince me otherwise, but yeah..it definitely was a nerf.

 

[Ijigen]

It's certainly a nerf. Beep-boop was a nice 50/50 at a huge % range or rarely a true combo into a kill sometimes.

I think the optimal throw game now might be upthrow until you want it out of your stale moves queue to kill (which is regrettably also a higher % in this game) and then backthrow until upthrow starts killing. That's 14.4% then 12%, and upthrow can combo at low percents as a bonus.

However, I've still been using down throw just as a curiosity and since it can be a useful tool if your opponent can't mash effectively. I recommend throwing it out at least once to gauge your opponent's mash capability. It would be interesting to see where top players (or at least top mashers) stack up in the spectrum between no mash and perfect mash. Are they 90% of the way there, or some other number? From there you could start talking about the realistic frame advantage and followups ROB will be getting, if any. For example I can often kill off dthrow->uair or dthrow->usmash in the ~40% band before I think upthrow will kill, but it's clearly not guaranteed in the theoretical. (also uair loses to waiting in bury, adding more uncertainty there. With a read of how and how quickly the opponent escapes you can hit it decently often though)

 

[EGsmash]

What's a real shame is that if your opponent doesn't know how to mash ROB actually has a chain grab on a lot of characters by doing 2x Dtilt and then re-grab, and that works all the way up to around 100%. Dtilt actually prevents the popup from happening so if you are able to hit one of them then it can really mess with your opponent. Unfortunately none of it works if your opponent knows how to push buttons. If you try anything fancy out of Dthrow then you just as easily end up looking stupid or opening yourself up to a brutal punish, versus the relatively safe 50/50 combo or kill that Beep-boop used to do.

 

[Gotmilk0112]

Most of us already know it's terrible. It only combos into up air / up smash past 120% or so. And by that point I'd probably just tack on a few more % and get a confirmed kill from up throw anyway.

Just left scratching my head at why the character with the shortest grab range in the entire game has the worst throws in the entire game.

 

[Luigifan18]

Most of us already know it's terrible. It only combos into up air / up smash past 120% or so. And by that point I'd probably just tack on a few more % and get a confirmed kill from up throw anyway.

Just left scratching my head at why the character with the shortest grab range in the entire game has the worst throws in the entire game.

Uh... are you sure R.O.B.'s grab range is shorter than the puffballs? Kirby and Jigglypuff have tiny arms!

 

[Gotmilk0112]

Uh... are you sure R.O.B.'s grab range is shorter than the puffballs? Kirby and Jigglypuff have tiny arms!

Maybe it was changed for Ultimate, but I remember in Smash 4 someone actually tested the frame data and found that ROB had the shortest grab range in the game (by a decent margin too)