Q&A Mechanics & Techniques Discussion

[Dr. Tuen]

Woo, the knowledge is real!

Thanks all. Effectively, after my dissertation is done this term, I'm going to use my newfound free time (see: unemployment) to write a Smash 4 Competitive Handbook. It'll hopefully reduce the number of repeat questions that get asked here... you know, like the one I just asked, ha ha.

Speaking of repeat questions...

Are the formula for shield hit lag and regular hit lag the same? The ZSS boards are discussing frame synced nairs, and I realized that outside of moves that generate landing hitboxes, frame syncs only work on hit... and I don't know how hit lag works on hit.

 

[A_Kae]

Woo, the knowledge is real!

Thanks all. Effectively, after my dissertation is done this term, I'm going to use my newfound free time (see: unemployment) to write a Smash 4 Competitive Handbook. It'll hopefully reduce the number of repeat questions that get asked here... you know, like the one I just asked, ha ha.

Speaking of repeat questions...

Are the formula for shield hit lag and regular hit lag the same? The ZSS boards are discussing frame synced nairs, and I realized that outside of moves that generate landing hitboxes, frame syncs only work on hit... and I don't know how hit lag works on hit.

Not quite. Hitlag on shield is exactly the same as on hit except for when the hitlag modifer is greater than 1. If that's the case, the modifer is divided by 1.25, to a minimum of 1. So, 1.8 becomes 1.44, and 1.1 becomes 1. Note that the division only occurs for regular modifers. Electric attacks aren't affected (unless they have a regular hitlag mod). Otherwise, they're identical.

Also, hitlag formula for completion's sake. (((d/2.6)+5)*h) d is the damage of the attack, h is the hitlag modifer.

Edit: Just in case it wasn't clear, it's hitlag on shield that divided in the case of modifiers greater than 1.

 

[Dr. Tuen]

Not quite. Hitlag on shield is exactly the same as on hit except for when the hitlag modifer is greater than 1. If that's the case, the modifer is divided by 1.25, to a minimum of 1. So, 1.8 becomes 1.44, and 1.1 becomes 1. Note that the division only occurs for regular modifers. Electric attacks aren't affected (unless they have a regular hitlag mod). Otherwise, they're identical.

Also, hitlag formula for completion's sake. (((d/2.6)+5)*h) d is the damage of the attack, h is the hitlag modifer.

Edit: Just in case it wasn't clear, it's hitlag on shield that divided in the case of modifiers greater than 1.

Radical. So, on-hit hit lag follows the equation to the T for all purposes, including modifiers over 1.0. Hitlag on shield with modifiers greater than 1 are divided by 1.25, but are not lower than 1 as a result. Electric attacks on shield are unaffected.

Can an attack have electric property AND a modifier? Lile... Electric + 1.5? How does that stack, if it even stacks at all, do we know?

 

[Lavani]

Can an attack have electric property AND a modifier? Lile... Electric + 1.5? How does that stack, if it even stacks at all, do we know?

Electric is x1.5 to the end result, it does indeed stack.

 

[A_Kae]

Radical. So, on-hit hit lag follows the equation to the T for all purposes, including modifiers over 1.0. Hitlag on shield with modifiers greater than 1 are divided by 1.25, but are not lower than 1 as a result. Electric attacks on shield are unaffected.

Can an attack have electric property AND a modifier? Lile... Electric + 1.5? How does that stack, if it even stacks at all, do we know?

Yep. The electric effect is basically just an additional 1.5x modifer. If something already had a 1.5x mod, electric would make that (effectively) 2.25 on hit, and 1.8 on shield.

 

[Dr. Tuen]

Excellent. Thanks for all the clarification! I'll add this to my list of referenced posts for sure!

 

[jmjb]

who has the best/safest side step in the game? like if you were to repeated spam side step in one spot, who has the best chance of dodging an attack based solely on spamming in place? lol weird question but I'm trying to try something silly, thanks.

 

[Vipermoon]

who has the best/safest side step in the game? like if you were to repeated spam side step in one spot, who has the best chance of dodging an attack based solely on spamming in place? lol weird question but I'm trying to try something silly, thanks.

I don't think there is a database about this so you'll have to check out each character separately on http://kuroganehammer.com

Pay attention to the start-up (time it takes for invincibility), the duration of it, and end lag.

 

[Lavani]

I don't think there is a database about this so you'll have to check out each character separately on http://kuroganehammer.com

Pay attention to the start-up (time it takes for invincibility), the duration of it, and end lag.

http://smashboards.com/threads/roll-air-dodge-frame-data.401323/

Although Z-axis displacement is also a thing with spot dodges that would impact safety, so it's not something that can be determined solely by frame data.

EDIT: And...I forgot that multiple patches have nerfed dodges since I posted that thread, so it's missing Lucas/Roy/Ryu and the numbers aren't entirely accurate anymore. I'll update it at some point tonight I guess.

EDIT2: Took less time than I thought, it's updated♫

 

[san.]

What frames does powershield drop again? I know shield drop is gone and shield stun is still there. What about attacker and defender shield hitlag?

 

[LancerStaff]

Maybe I'm just stupid, but they miiiight of removed vectoring completely. Pit's Uthrow kills at 198% on FD against Mario regardless of wether or not I hold up or down.

...Hope I'm wrong. Don't wanna re-re-relearn DI.

 

[Rawbinator]

^Pretty sure it's been that way for quite a while now. If the knockback is vertical it's better to hold left or right
See this vid at 3 minutes:
https://www.youtube.com/watch?v=PnqgfP7vrTs

 

[[Deuce]]

Let me make sure I have this straight-

This patch, shield hitlag was changed so that it is affected by the attacker's hitlag modifier, whereas it was always 1.0x before?

So for example any of Marth's sourspot moves before were 0.7x modifier and the defender experienced 1.0x, but now its 0.7x attacker /0.7x defender

and for electric moves which had a modifier of 1.5 (ie. >1), the defender used to experience 1.0x but now its 1.5÷1.25 = 1.2x for the defender, meaning 1.5x attacker/1.2x defender, correct?

So to extrapolate that info, ZSS got sidegraded or maybe a tiny upgrade as her paralyzer and D-smash induce maybe less shield hitlag(projectile but electric) but zair and nair (electric non projectile) induce more than before?

 

[Lavani]

Let me make sure I have this straight-

This patch, shield hitlag was changed so that it is affected by the attacker's hitlag modifier, whereas it was always 1.0x before?

So for example any of Marth's sourspot moves before were 0.7x modifier and the defender experienced 1.0x, but now its 0.7x attacker /0.7x defender

and for electric moves which had a modifier of 1.5 (ie. >1), the defender used to experience 1.0x but now its 1.5÷1.25 = 1.2x for the defender, meaning 1.5x attacker/1.2x defender, correct?

So to extrapolate that info, ZSS got sidegraded or maybe a tiny upgrade as her paralyzer and D-smash induce maybe less shield hitlag(projectile but electric) but zair and nair (electric non projectile) induce more than before?

Electric's extra hitlag isn't divided, and if it were it'd be divided for both characters.

1.0 hitlag electric: used to be 1.5x/1.0x, now 1.5x/1.5x
1.5 hitlag: used to be 1.2x/1.0x, now 1.2x/1.2x

Maybe I'm just stupid, but they miiiight of removed vectoring completely. Pit's Uthrow kills at 198% on FD against Mario regardless of wether or not I hold up or down.

...Hope I'm wrong. Don't wanna re-re-relearn DI.

It's been this way since 1.0.4 3DS. Holding up/down doesn't alter knockback against vertical hits (within 80°~100°, I believe), but still does otherwise.

 

[Big O]

What frames does powershield drop again? I know shield drop is gone and shield stun is still there. What about attacker and defender shield hitlag?

Powershielding has it's own separate (usually smaller) shieldstun value and the 7 frame shield drop animation gets special non-movement based IASA. Basically, powershielding decreases shieldstun by a small amount and lets you cancel the shield drop lag with attacks and specials in addition to the normal OoS options (jump, roll/dodge, grab). You can't cancel the shield drop lag with dashing, walking, crouching, or shielding. As far as I can tell, there is no minimum shield duration (commonly referred to as the 11 frames of shield lock) for powershielded hits.

 

[LancerStaff]

It's been this way since 1.0.4 3DS. Holding up/down doesn't alter knockback against vertical hits (within 80°~100°, I believe), but still does otherwise.

So you can still vector but only against more angled hits, right?

 

[Lavani]

So you can still vector but only against more angled hits, right?

Right. Vertical launches are standard DI rules, otherwise knockback is increased/decreased by DIing up/down.

 

[Horseketchup]

[(Damage / 3.5) + 2] rounded down

^ So has this value for projectile shieldstun been 100% confirmed? Saw some posts a weeks back about a reliable Japanese player who discovered it postpatch, didn't see any further testing or confirmation about it though.

Also just curious, does there happen to be a list of all the moves in the game with hitlag modifiers?

 

[Lavani]

Also just curious, does there happen to be a list of all the moves in the game with hitlag modifiers?

There's a Japanese table by the same guy you mentioned regarding your first question. For lazy English reference I'm not sure if we have anything better than Thinkaman's 1.0.4 pastebin, which is serviceable aside from lacking DLC characters (and hitlag changes over patches, if there were any?) for obvious reasons, and I don't think is really the type of format you're looking for.

 

[Horseketchup]

Thanks, well at least that pastebin should be a pretty good reference. Yea I've seen that table, but it's just the only testing or info I've seen, just was hoping maybe there were others who have looked into it too.

EDIT: Actually that table isnt the one i was thinking of, this looks really detailed but even with autotranslate I can't quite understand everything

 

[Lavani]

It seems like if Ike clanks a projectile with his dash attack he can connect both the 14% strong hit and one of the weak hits on the same opponent. Dunno if it's unique to his dash attack or if it can be reproduced by other attacks with similar attributes (Isn't canceled after clank? Multiple hitboxes? Early/late hits?), but it's pretty interesting regardless.

EDIT: Actually that table isnt the one i was thinking of, this looks really detailed but even with autotranslate I can't quite understand everything

Oops, didn't see the edit until now.

Character names are on the Y axis, in 3DS roster order but with Luma separate from and after Rosalina. X axis has attacks in the following order (no special moves):

Jab1 > Jab2 > Jab3 > Rapid Jab Ender > Dash Attack > Ftilt > Utilt > Dtilt > Fsmash > Usmash > Dsmash > Nair > Fair > Bair > Uair > Dair

 

[LancerStaff]

That table's pretty cool. But is there one for specials?

 

[san.]

It seems like if Ike clanks a projectile with his dash attack he can connect both the 14% strong hit and one of the weak hits on the same opponent. Dunno if it's unique to his dash attack or if it can be reproduced by other attacks with similar attributes (Isn't canceled after clank? Multiple hitboxes? Early/late hits?), but it's pretty interesting regardless.


Oops, didn't see the edit until now.

Character names are on the Y axis, in 3DS roster order but with Luma separate from and after Rosalina. X axis has attacks in the following order (no special moves):

Jab1 > Jab2 > Jab3 > Rapid Jab Ender > Dash Attack > Ftilt > Utilt > Dtilt > Fsmash > Usmash > Dsmash > Nair > Fair > Bair > Uair > Dair

I believe it can be reproduced, it's just easiest with Ike. The only other possible candidate... Jigglypuff's dash attack? I'm not sure.

 

[LancerStaff]

So, anybody mind running me through how to calculate safety on shield?

For example, Pit's Nair final hit is on frame 25, ACs on 30, does 4%, has a hitlag modifier of 1.8, and his soft and hard landings are 2 and 5 respectively.

 

[A_Kae]

So, anybody mind running me through how to calculate safety on shield?

For example, Pit's Nair final hit is on frame 25, ACs on 30, does 4%, has a hitlag modifier of 1.8, and his soft and hard landings are 2 and 5 respectively.

Shield Advantage = (s-(t-a))

h = Shieldstun Frames
t = Total Frames
a = Attack Frame

For the example that you provided, h would be 4, t would be 35 (assuming that you land on frame 30, 5 frames of landing lag would make 35), and a would be 25.

(4-(35-25)) = -6 (add 7 for shield drop)

E: Shieldstun is ((d/1.72)+2), where d is the attack's damage. Also, since 1.1.1, you don't need to take hitlag in to account for shield safety.

 

[LancerStaff]

Shield Advantage = (s-(t-a))

h = Shieldstun Frames
t = Total Frames
a = Attack Frame

For the example that you provided, h would be 4, t would be 35 (assuming that you land on frame 30, 5 frames of landing lag would make 35), and a would be 25.

(4-(35-25)) = -6 (add 7 for shield drop)

E: Shieldstun is ((d/1.72)+2), where d is the attack's damage. Also, since 1.1.1, you don't need to take hitlag in to account for shield safety.

So if Pit's Fair does 4% on the final hit it'd effectively do the same amount of stun even with a 2.2 hitlag modifier? If that's the case then it'd go (4-(33-18)) = -11 on a hard landing. You can get soft landings on ACs right?

 

[A_Kae]

So if Pit's Fair does 4% on the final hit it'd effectively do the same amount of stun even with a 2.2 hitlag modifier? If that's the case then it'd go (4-(33-18)) = -11 on a hard landing.

That all looks right to me, yes.

You can get soft landings on ACs right?

If the move animation is still in progress when you land it'll be hard landing lag. For example, Marth's shorthop is 41 frames, fair is 37, and uar is 45 (and both autocancel earlier than f40). Shorthopped fair will take soft landing, since the move is finished when you land, but uair is hard landing, since the move hasn't ended by the time that you land.

How many frames is Pit's shorthop?

 

[LancerStaff]

That all looks right to me, yes.



If the move animation is still in progress when you land it'll be hard landing lag. For example, Marth's shorthop is 41 frames, fair is 37, and uar is 45 (and both autocancel earlier than f40). Shorthopped fair will take soft landing, since the move is finished when you land, but uair is hard landing, since the move hasn't ended by the time that you land.

How many frames is Pit's shorthop?

Pit's SH I believe is about 38-40 frames. Kurogane doesn't have it listed but Dair lasts 37 frames and you can just barely jump out of it during a SH. So from the looks of things Dair and maybe Bair could end up with a soft landing, though Bair definitely wouldn't be worth it regardless.

Part of the reason I wanted to do this was that I wanted to double-check Shaya's data since I wasn't sure if he was assuming it was hard landing or soft landing. Actually, Shaya must've calculated it based on the soft landing if he said Nair was -3.

Oh yeah, when calculating the shieldstun, you always round down right?

 

[A_Kae]

Pit's SH I believe is about 38-40 frames. Kurogane doesn't have it listed but Dair lasts 37 frames and you can just barely jump out of it during a SH. So from the looks of things Dair and maybe Bair could end up with a soft landing, though Bair definitely wouldn't be worth it regardless.

Part of the reason I wanted to do this was that I wanted to double-check Shaya's data since I wasn't sure if he was assuming it was hard landing or soft landing. Actually, Shaya must've calculated it based on the soft landing if he said Nair was -3.

Oh yeah, when calculating the shieldstun, you always round down right?

I should have mentioned in my last post that it's entirely possible that I'm wrong about how soft/hard landing works. This is just what I've been assuming from my tests a long time ago, I don't have any sort of confirmation on that and I could be misremembering what I found.

Shieldstun is always rounded down, yes.

 

[LancerStaff]

I should have mentioned in my last post that it's entirely possible that I'm wrong about how soft/hard landing works. This is just what I've been assuming from my tests a long time ago, I don't have any sort of confirmation on that and I could be misremembering what I found.

Shieldstun is always rounded down, yes.

Well, I guess that's my next question for the thread. Because I came to a similar conclusion since I don't think I've gotten a soft landing off of SH anything besides Dair.

 

[Vipermoon]

A_Kae

I had no idea landing during autocancel frames with no fast fall could give hard landing lag. I tried this myself in 1/4 training and it was weird. I'm hoping someone with recording equipment could try this out...

Marth has a slow gravity so he doesn't reach top fall speed out of a short hop. When I short hop and land with shield there is barely a landing animation and it looks as though I can shield immediately (soft landing?). However, when I fall from a platform, full hop, fast fall, or land in an autocancel window I see a more typical landing animation and I can only shield noticeably later.

So we have to find out if landing at max fall speed gives hard landing lag just like it does for fast falling. We also have to make sure, like you guys were talking about, if autocancel windows are hard landing lag.

 

[LancerStaff]

A_Kae

I had no idea landing during autocancel frames with no fast fall could give hard landing lag. I tried this myself in 1/4 training and it was weird. I'm hoping someone with equipment could try this out...

Marth has a slow gravity so he doesn't reach top fall speed out of a short hop. When I short hop and land with shield there is barely a landing animation and it looks as though I can shield immediately (soft landing?). However, when I fall from a platform, full hop, fast fall, or land in an autocancel window I see a more typical landing animation and I can only shield noticeably later.

So we have to find out if landing at max fall speed gives hard landing lag just like it does for fast falling. We also have to make sure, like you guys were talking about, if autocancel windows are hard landing lag.

I have a freakin' ton of equipment. What exactly do you want me to do?

 

[Lavani]

I don't need to turn the game on to tell you that you don't need to be fastfalling for hard landing lag. I wouldn't know about the autocancel thing, though.

 

[Vipermoon]

I don't need to turn the game on to tell you that you don't need to be fastfalling for hard landing lag. I wouldn't know about the autocancel thing, though.

Oh okay so it is safe to assume that the squat landing animation always means hard landing lag? Also how do you know?

LancerStaff I wasn't talking about in-game equipment I meant recording equipment to count the landing frames in various situations.

 

[Dr. Tuen]

Shield Advantage = (s-(t-a))

h = Shieldstun Frames
t = Total Frames
a = Attack Frame

For the example that you provided, h would be 4, t would be 35 (assuming that you land on frame 30, 5 frames of landing lag would make 35), and a would be 25.

(4-(35-25)) = -6 (add 7 for shield drop)

E: Shieldstun is ((d/1.72)+2), where d is the attack's damage. Also, since 1.1.1, you don't need to take hitlag in to account for shield safety.

There's a correction to be made here.

There are two shield safety values. First, obtained here, is if there are no shield lock frames. Shield lock used to be a bigger deal, since it could add 7 frames to your in-shield time. In 1.1.1 they made it so shield lock frames overlap with shield stun frames. However, if your applied shield stun is less than 7 frames long, and you hit someone on the first non powershield frame (frame 4?), then you'll need to add shield lock frames.

In this case, your shield stun frames are 4, and total potential shield lock is 7. So, 7-4=3 shield lock frames.

Without shield lock: -6, +1 after shield drop
With max shield lock: -9, -2 after shield drop

 

[Lavani]

Oh okay so it is safe to assume that the squat landing animation always means hard landing lag? Also how do you know?

Full jumps having a bit more landing lag than shorthops was something I noticed when I was trying to eyeball Megaman and Greninja's jumpsquats around release, and it was that way in previous games too I believe. You're right about the animation being different between the two.

I don't know if you actually need to be at your max normal fall speed for a hard landing or if it can happen earlier than that though, that was something I didn't get to look into before my 3DS broke.

 

[A_Kae]

There's a correction to be made here.

There are two shield safety values. First, obtained here, is if there are no shield lock frames. Shield lock used to be a bigger deal, since it could add 7 frames to your in-shield time. In 1.1.1 they made it so shield lock frames overlap with shield stun frames. However, if your applied shield stun is less than 7 frames long, and you hit someone on the first non powershield frame (frame 4?), then you'll need to add shield lock frames.

In this case, your shield stun frames are 4, and total potential shield lock is 7. So, 7-4=3 shield lock frames.

Without shield lock: -6, +1 after shield drop
With max shield lock: -9, -2 after shield drop

You're absolutely right, of course. I just prefer to ignore shield lock for my calculations since it's not something that's going to be consistent (you're not always going to hit on f4 of a shield), and it doesn't apply to OoS stuff.

Edit: Actually. it should be -3/+4 with max shield lock.

 

[LancerStaff]

LancerStaff I wasn't talking about in-game equipment I meant recording equipment to count the landing frames in various situations.

Nope, nothing like that. Besides uploading replays... But that's 30fps and drops frames or something.

You're absolutely right, of course. I just prefer to ignore shield lock for my calculations since it's not something that's going to be consistent (you're not always going to hit on f4 of a shield), and it doesn't apply to OoS stuff.

Edit: Actually. it should be -3/+4 with max shield lock.

So how much can safety fluctuate from this mechanic?

Well actually I don't think it'd make a difference on Nair and Fair because they're multihit but still curious.

 

[A_Kae]

So how much can safety fluctuate from this mechanic?

Well actually I don't think it'd make a difference on Nair and Fair because they're multihit but still curious.

Well, there's 7 frames of shield lock, and a minimum 2 frames of shield stun, so at most it's 5 extra frames of safety.

 

[Dr. Tuen]

You're absolutely right, of course. I just prefer to ignore shield lock for my calculations since it's not something that's going to be consistent (you're not always going to hit on f4 of a shield), and it doesn't apply to OoS stuff.

Edit: Actually. it should be -3/+4 with max shield lock.

Ah, my error there may be due to my mixing of + and - for shield advantage. I'm very, very, dyslexic, so many choices are fine, but deciding between two is just obnoxiously hard to remember.

So, for clarification... how is this set up? Is more negative more safe?