Link's Chaingrab And When It Works

[EmptySky00]

Before I get into the raw data, I would like to first give an explanation of the dynamics of the chain grab.
Links dthrow as of 3.6 has a release point that is slightly behind him. This means that on no DI you need to do a turnaround grab. This requires you hold back for 1 frame before grabbing. The changed release point also means that DI back is more effective against his dthrow. DI back will almost always send them back enough that you will need to dash grab, which including the 2 frames for dash to begin puts it at frame 14. Characters also hit the ground slightly faster if they DI to the side, so that also makes the timing more strict. Note that dash grab will miss no DI and turn around grab will miss DI back. The opponent can also DI in front of you to throw you off further. Basically, at the lowest possible percent the chain grab works and for some time after you need perfect reaction and frame precision. This is even further complicated by the fact that Dthrow is weight dependent which means the perfect timing for everyone is going to be different.
DI effectively extends how long the chain grab lasts. So if you're being CG'd and are sure it no longer works, it might be best to go with no DI.

Another helpful resource involving Dthrow: http://smashboards.com/threads/the-empty-combo-links-dthrow-followups-and-when-they-work.401485/
This was made in 3.5 but the ranges for everyone who didn't have their weight changed such as Yoshi and ROB are exactly the same.

 

[Deleted member]

can you give me a % range of when its guaranteed on any DI? i think thats a lot more useful for a real player

 

[EmptySky00]

Done. Denoting characters it never works on right now.

Edit: double done. A dash denotes that it doesn't work or the information isn't relevant. It can be considered to work on any DI when it starts working on DI back, so I stopped finding the first number. It also doesn't work on the super heavyweights. They can just jump out on no DI.

 

[Deleted member]

 

[smokebomb12]

Impressive.

 

[EmptySky00]

I'm going to do spacies and **** just for the sake of completion. Also if you edit those in, I think sonic should say 65 and ganondorf 105 just for the sake of uniformity. I rounded down to the nearest 5 with everyone else as I adapted my methodology more.

Edit: As a real note I think it should be designated when you have to start pivoting to continue the chain grab.

 

[JesteRace]

Wow. This is cool. Very useful information.

 

[Thor]

I'm going to do spacies and **** just for the sake of completion. Also if you edit those in, I think sonic should say 65 and ganondorf 105 just for the sake of uniformity. I rounded down to the nearest 5 with everyone else as I adapted my methodology more.

Edit: As a real note I think it should be designated when you have to start pivoting to continue the chain grab.

I prefer exact percents where possible [especially since it's already in there], but if you do round stuff, at least make an asterisk at the top of the chart saying that some stuff is rounded down, so people don't come in and complain that they can CG someone at some percent outside the range listed.

 

[EmptySky00]

I really couldn't care less if people complain about something so minor. They didn't take the time to compile the data so they can live with a 2% discrepancy. Exact percents will seldom come into play nor will it matter when they do as long as I don't overshoot the end %. Real play isn't exact. Even at the escape %s they often have around 1 frame to jump out.

 

[E2xD]

Have you tried U-throw CGs (on spacies and such)

 

[EmptySky00]

No, I haven't. I could try to test it next time I'm hardcore labbing. But I have no idea when those work. It's definitely not at 0.

As another note, these chaingrabs can go on longer than denoted if they keep DI'ing to the side lol. So you can **** with the ignorant and just react to the no DI input with up B or something.

 

[Kaeldiar]

How has no one pointed out that you wrote this information IN A DEATH NOTE????

 

[Titanium]

Hey bud, not trying to be a ****, but you should check your numbers again. I've been playing around with regrabs myself, and I think you're using when the opponent's body would be within a grab-able height, but aren't accounting for them jumping out. The frame advantage at 0% is:

- Jigglypuff: 16
- Fox/G&W: 12
- Kirby/Meta Knight: 11
- Falco/Pikachu/Lucas/Pit/Squirtle/Sonic: 10
- Diddy/Ivysaur/Roy/ZSS/Toon Link/Wolf/Marth: 9
- IC/Peach/Zelda/Sheik/Olimar: 8
- Ness/Mewtwo: 7
- Mario/Lucario/Luigi/C Falcon/Link: 6
- Snake/Ike/ROB/Yoshi: 5
- Wario/Ganon/Samus/Charizard/DDD: 4
- DK: 3
- Bowser: 2

You gain 1 frame of advantage for ever so slightly less than each additional 12% (5% is the starting point, in case you're curious). For the regrab to be true for all DIs, you need to have a minimum of 12 frames of advantage so that you can turn around and JC/dash grab. This will make your handbox active for the first frame they can jump and show you if the chaingrab is true or not. I'll post the data below as to when the regrab is true (they can't jump out of it):

G&W: 0 – 67 (must start pivot grabbing no DI at 67)
Pit: 4 – 9; 17 - 55 (not true early on, but he has no response, must be JC grabbed for DI away at low percents or he can hit the ground)
Kirby: 4 – 19
Squirtle: 15 - 83 (must start pivot grabbing no DI at 73)
Pikachu: 17 – 88 (must start pivot grabbing no DI at 81)
Sonic: 20 – 93 (must pivot grab no DI at 93)
MK: 50 – 178 (must start pivot grabbing no DI at 166)
Tink: 60 - 139 (must start pivot grabbing no DI at 119)
Lucas: 62 – 141 (must start pivot grabbing no DI at 141)
ZSS: 65 - 132 (must start pivot grabbing no DI at 129)
Sheik: 73 – 136 (must start pivot grabbing no DI at 129)
Roy: 75 - 163 (must start pivot grabbing no DI at 144)
Diddy: 76 – 175
Lucario: 89 – 119 (must start pivot grabbing no DI at 107)
DDD: 77 – 112 (not true at low percents but he has no response)
Ganon: 112 - 112…I know, right?
DK: 112 - 113

If a character is not on the list above, they cannot be chaingrabbed by Link, at least not in any way that's true. If the character is on the list above, all the data applies to all DI. The most relevant info is the 73-death (dair is confirmed) on Sheik and the 50-death on MK. We also have a confirmed tech chase on MK from 0-60 so if you get a grab on him, there is no reason he shouldn't lose a stock.

PS: all the percents above are the damage after a grab so staling does not affect the validity.
PPS: for chaingrabbing DDD and Pit at low percents, their jump doesn't immediately increase their height and they don't have fast enough aerials to harm Link (including their up-b's) so the chain grab is not technically a combo, but it is still true.

- Titanium <3

 

[EmptySky00]

Hey bud, not trying to be a ****, but you should check your numbers again. I've been playing around with regrabs myself, and I think you're using when the opponent's body would be within a grab-able height, but aren't accounting for them jumping out. The frame advantage at 0% is:

- Jigglypuff: 16
- Fox/G&W: 12
- Kirby/Meta Knight: 11
- Falco/Pikachu/Lucas/Pit/Squirtle/Sonic: 10
- Diddy/Ivysaur/Roy/ZSS/Toon Link/Wolf/Marth: 9
- IC/Peach/Zelda/Sheik/Olimar: 8
- Ness/Mewtwo: 7
- Mario/Lucario/Luigi/C Falcon/Link: 6
- Snake/Ike/ROB/Yoshi: 5
- Wario/Ganon/Samus/Charizard/DDD: 4
- DK: 3
- Bowser: 2

You gain 1 frame of advantage for ever so slightly less than each additional 12% (5% is the starting point, in case you're curious). For the regrab to be true for all DIs, you need to have a minimum of 12 frames of advantage so that you can turn around and JC/dash grab. This will make your handbox active for the first frame they can jump and show you if the chaingrab is true or not. I'll post the data below as to when the regrab is true (they can't jump out of it):

G&W: 0 – 67 (must start pivot grabbing no DI at 67)
Pit: 4 – 9; 17 - 55 (not true early on, but he has no response, must be JC grabbed for DI away at low percents or he can hit the ground)
Kirby: 4 – 19
Squirtle: 15 - 83 (must start pivot grabbing no DI at 73)
Pikachu: 17 – 88 (must start pivot grabbing no DI at 81)
Sonic: 20 – 93 (must pivot grab no DI at 93)
MK: 50 – 178 (must start pivot grabbing no DI at 166)
Tink: 60 - 139 (must start pivot grabbing no DI at 119)
Lucas: 62 – 141 (must start pivot grabbing no DI at 141)
ZSS: 65 - 132 (must start pivot grabbing no DI at 129)
Sheik: 73 – 136 (must start pivot grabbing no DI at 129)
Roy: 75 - 163 (must start pivot grabbing no DI at 144)
Diddy: 76 – 175
Lucario: 89 – 119 (must start pivot grabbing no DI at 107)
DDD: 77 – 112 (not true at low percents but he has no response)
Ganon: 112 - 112…I know, right?
DK: 112 - 113

If a character is not on the list above, they cannot be chaingrabbed by Link, at least not in any way that's true. If the character is on the list above, all the data applies to all DI. The most relevant info is the 73-death (dair is confirmed) on Sheik and the 50-death on MK. We also have a confirmed tech chase on MK from 0-60 so if you get a grab on him, there is no reason he shouldn't lose a stock.

PS: all the percents above are the damage after a grab so staling does not affect the validity.
PPS: for chaingrabbing DDD and Pit at low percents, their jump doesn't immediately increase their height and they don't have fast enough aerials to harm Link (including their up-b's) so the chain grab is not technically a combo, but it is still true.

- Titanium <3

I looked at when they escaped stun. There were only a few that were dubious such as Ivysaur (which isn't worth using anyway because it lasts for like 12%) where she could barely Nair you and I denoted the beginning of the grab as when you would grab armor Nair. Aside from that I was using frame-by-frame and noting when they got out of stun, and thus when they could jump out. Virtually every regrab (minus Ivysaur as the only exception I can think of) was achieved while they were still in stun. On DI back, I took into consideration when they got out of stun along with when I could actually grab them on a frame perfect dash grab before they hit the ground. Then for testing the end %'s, I did no DI because that would be when they could jump out the earliest because of them going higher. If they DI back or forward it effectively lowers their launch height, and you'll basically always have the frame advantage to regrab after a certain point and the main factor will be launch height. You can obviously start CG'ing them earlier on no DI for this reason, which is why my numbers might appear odd to you, but I definitely thoroughly checked it. Umbreon suggested that I should only look for when it's absolutely guaranteed on correct DI, which could make it a tad misleading when the opponent isn't doing the absolute correct counterplay, but ya. The numbers are right in like 98% of instances and there was no real variable I could have ignored to my knowledge. I also rounded to the nearest 5 on the end %'s, because the CGs are inexact anyway. A 3% discrepancy doesn't mean anything. Just pummel once to make up for it.

Perfect dash grab is frame 14 btw. It takes 2 frames for a dash to register, then the dash grab itself is 12 frames. I made sure I was testing perfect inputs as well.

I'm curious as to how you're getting 170+% on meta knight, because I definitely saw it stop working at 150%. I'd be happy to compare measurement methodology with you to see if we can get more accurate numbers.


Edit: I also know all of the Dthrow stun trends for each character because I charted it out for everyone up to 150 ages ago. I also had it next to me while testing. Stun increases by 4 for every 50%. Bowser is 2 - 6 - 10 - 14, etc.

 

[Titanium]

I'll have to run it through again to ensure how many frames of what action you're doing with MK, but if memory serves at the super high percents (for no DI) you spend 8 frames not moving, 4 frames dashing back, hit grab and pivot during the second frame. You're right about DI back being EZ money at higher percents, but for a lot of the earlier percents, they want to jump out instead of hitting the ground. Take Ike for example, we have +10 on him if we d-throw him at 60% but on both DI back and no DI, he can input a frame perfect jump to escape. He can't hit us with a nair and he won't hit the ground, but there's no reason he would go for those options when jumping out is a perfect escape. I'm wondering if you're inputting the jump a frame late, since your numbers started working when I was jumping a frame later.

Yeah, d-throw stun does increase effectively at that rate, but if they're not in hitstun while your grabbox is out, I just think you need to ensure they can't jump out.

PS: there are situations where going for jump doesn't apply of course, I realize. G-dorf for example will be caught whether he jumps or not at 80% for no DI, but with DI back he can jump out.

 

[EmptySky00]

I'll have to run it through again to ensure how many frames of what action you're doing with MK, but if memory serves at the super high percents (for no DI) you spend 8 frames not moving, 4 frames dashing back, hit grab and pivot during the second frame. You're right about DI back being EZ money at higher percents, but for a lot of the earlier percents, they want to jump out instead of hitting the ground. Take Ike for example, we have +10 on him if we d-throw him at 60% but on both DI back and no DI, he can input a frame perfect jump to escape. He can't hit us with a nair and he won't hit the ground, but there's no reason he would go for those options when jumping out is a perfect escape. I'm wondering if you're inputting the jump a frame late, since your numbers started working when I was jumping a frame later.

Yeah, d-throw stun does increase effectively at that rate, but if they're not in hitstun while your grabbox is out, I just think you need to ensure they can't jump out.

PS: there are situations where going for jump doesn't apply of course, I realize. G-dorf for example will be caught whether he jumps or not at 80% for no DI, but with DI back he can jump out.

9/11 was bad.


As I said, they were effectively always in stun when the grab landed except for one irrelevant example that I can think of.

On DI back they generally hit the ground at the earliest possible grab percents before being out of stun.

 

[Titanium]

9/11 was bad.


As I said, they were effectively always in stun when the grab landed except for one irrelevant example that I can think of.

On DI back they generally hit the ground at the earliest possible grab percents before being out of stun.

Just try it, humour me. Mathematically (using the hitstun data) the heavies like Ike and up are out of hitstun for at least a frame before you grab them at 60.

 

[EmptySky00]

How has no one pointed out that you wrote this information IN A DEATH NOTE????

Intentional hehe.


Ike may have been one of the iffy ones while I was still adapting measurements. I'll look into it. Any that exist would have only been for a frame. So I'll see if his jump is fast enough. Even if it's not fast enough on DI back it may be on no DI. but he's one of the very few exceptions and our data still differs significantly so there's obviously more to it.

 

[EmptySky00]

Ya it doesn't work on Ike apparently. The grab box doesn't interact with his leg. So you got me on that one. Any other discrepancies you noted?

 

[Titanium]

Well, given the big differences in our results, a lot. Add me on skype "herooftiming" and we can discuss it. Hopefully figure out what's going on and post a comprehensive list for the Link community.

 

[EmptySky00]

I'm pretty sure I already know what it is tbh. The ike thing happened before I was paying attention to hitstun, I'll admit it. And the rest is something else. I'm just incompetent at this point don't mind me. I've never really compiled data for something with more than like 2 variables before and I just overlooked things. So you're right I'm stupid. etc.

 

[Deleted member]

send me the updated info and i'll fix the excel picture for it

 

[EmptySky00]

I'll look into it at some point.