One of the biggest things we’ve all experienced is how inconsistent RT and sword loops can get, so I want to devote this thread to finding the most consistent and optimized versions and share some of the notes I’ve been gathering. So please go over these and offer anything you can. I’m sure we’ve all had numerous cases of “random” drops in vergil’s combos and I’m beginning to piece together the rational reasoning of why.
-----
Current finds and key notes
[list]
[]Vergil’s sword loops (or dpf+:m:…but not likely) have side specific properties - I discovered this recently and it was both infuriating to find and relieving to finally have a answers to combos dropping. A simple test to see the effects, do: c.:h:, activate swords, dpf+:m: on each player side with a normal body type. On player 2 side the feet will be right at vergil’s face nearly on top of him, on player 1 side the feet will be chest height and farther forward. Another test is in XF3 do: c.:h:, activate swords, dpf+:m:, c.:h:, dpf+:m:. On player 1 side it will always connect, on player 2 side they will always flip out.
[]Different characters will float differently with different loops - This is no real secret by any means, but we’ve pretty much ruled the standard c.:h:, qcf+:l:, c.:h: is doomed to be a relic given how easy it is for characters to float to high before the first c.:h:, drop out of bubble too quick and fall out of the 2nd c.:h:, or simply float too far out. We are working on ways to be more consistent across the board.
[]:s::h: does suffer hit stun deterioration, but very late in combos - This has lead to numerous improvements in combos and is the key to a lot of new tech to be shown later.
[]qcf+:atk: has very random juggle properties - This move while still very much a key role in loops is the usual culprit in inconsistencies. It can push them too far or drop them sooner than expected. There are steps to dissuade these anomalies however.
[]c.:h:, RT, qcf+:atk:, dpf+:m:, j.d+:h: df+:h:, j.:s:, and swords are the only moves immune to hit stun deterioration for the full extent of the move - These moves will not change how they affect an opponent early or late into combos. Moves that do, can sometimes be useful by immediately hitting into these moves (EX: RT, c.:h::s:, qcf+:l: (TK cancel). Normally :s: actually will scale late (:s::h: will eventually have a gap between), but it connects just before RT hits, setting up a consistent qcf+:l: every time)
[]crouch and standing effect RT and swords - You can alter combos by crouching or standing and it will lower both of these to create new possibilities or risks of dropping. s.:h:, :s:, :s::h:, qcf+:atk:, and dpf motion, dashing and teleporting also create a brief moment of crouch.
[]j.d+:h: won’t hardknock down if Vergil is below the target when swords are fully active - Swords eventually produces a gap as swords disappear one by one allowing a space for the opponent to fall, be aware of this.
[]s.
, s.:h:, :s:, :s::h:, RT all move Vergil forward, qcf+:atk: moves him backwards - Keep this in mind as these can be used to improve consistency or push you the wrong way.
[*]Swords disappear the more they hit - Many moves help prevent swords from wasting hits early and allow for more time before they disappear. This can drastically effect when you need to throw RT. Happy birthdays, X-factor, DT can all cause the swords to run out much quicker than normal. They will also disappear one by one, leaving a gap.
[/list]
-----
That said…I have a listing of several RT/sword loops with notes on consistency benefits and common problems…
RT Loops
[list]
[*]RT, , :s::h:, qcf+:l:, s.:h:/c.:h: ( = slight pause, crouch pause, or walk forward briefly…you must do one of these) - this is proving to be the best RT loop for both damage, meter, and helping to recharge RT in combos. A great case of showing the recharge benefit is doing a wall bounce RT set up, use this loop, then use s.:h::s::h: (ground bounce) ~RT into this loop again. Without that last hit RT would not be charged. This opens up huge assist possibilities too. There are oddities though. If you do a slight pause in place during " * ", timing s.:h:/c.:h: becomes a little tight. s.:h: is worth trying, but c.:h: is VERY risky and only worth doing on floaty characters like Zero. If you hold forward during " * " it is VERY consistent and they will bounce higher on the bubble and closer to you, you will ALWAYS be able to land c.:h: or s.:h:…but they can bounce too high causing c.:h: to whiff if too early or s.:h: to have them too high for a favorable follow up; also note to let go before hitting :s: or your bubble will become a dpf+:l:. If you hold down during " * " it seems to slightly improve the :s::h: confirm to avoid flipping them out and bringing them slightly closer to you, I assume this is because RT is lowers slightly as it returns. All around the big issue is simply making that last hit after bubble consistent for combos as it has unique potential that the other loops can’t provide with that RT charge.
For swords this is VERY reliable for making sure the opponent is next to you when you activate. The main issue here is finding the most consistent and rewarding Sword loop. immediate dpf+:m: will flip past them in most cases. The most consistent method so far is to omit the last hit (which will be c.:h: if going to swords), due the walk forward during " * ", then immediately back dash (hold down ASAP), then dpf+:m:, dash (side switch), RT ,etc… That dash can be a bit complicated.
[*]RT, c.:h:, qcf+:l:, c.:h: - The classic. Great because it ignores hitstun scalling fully…bad because it is notorious for whiffing the c.:h: on certain characters or pushing the opponent too far to combo. It works great when they are flush next to you and a little high, but still has some random factors.
[*]RT, s.:h:, qcf+:l: - This is great early in combos when s.:h: won’t flip out, but it’s very consistent and lets you link off the bubble.
[*]RT, c.:h::s:, qcf+:l: (TK cancel) - The timing on this is fast, but so far works great. This is only useful if immediately going into swords during bubble. In some RT set ups, you run the risk of c.:h: whiffing on characters like Modok if mistimed.
[*]RT, delay, s.:h:, qcb+:l: - Simple, easy early in combos, and moves you forward for consistency. Has low reward though and limited use.
[*]RT, c.:h:, c.:h: - super consistent for setting up sword loops…but low reward and the same c.:h: whiff possibility in high RT set ups.
[/list]
-----
Sword Loops
[list]
[]Swords, dpf+:m:, dash (optional side switch), RT, RT loop… - Your normal late combo sword loop. The only issue is that the dpf+:m: is really picky in a lot of scenarios. Many Sword loop set ups can easily have it too high, too close, too far, etc…and don’t for get the side specific issue ( D: ). Ideally this is what you want. You can do adjustments like crouching briefly, teleporting, walking forewards/backwards or dash (hold down ASAP), etc… but those adjustments can often risk valuable time which can deprive you of the sideswitch (which nets more meter and can improve follow up consistency dramatically). I have yet to find a universal answer for all set ups and scenarios.
[]Swords, j.d+:h:, qcb+:m:, crouch, :s::h:~RT, RT loops - Seems to work fine sometimes, but :s::h: gets riskier in long combos once swords are involved and you can risk hitting them out.
[]Swords, dash. j.d+:h:, dash, j.d+:h:, RT, RT loops - Good for salvaging a far opponent from dropping, but can risk unwanted hard knockdowns and the RT can have some variance because swords may still slightly be active.
[]Swords, dpf+:l:~:h:, dash j.d+:h:, RT, RT loops - Works fine at first, but can occasionally kick them out of loops or hardknockdown. You also risk miss timing the teleport.
[*]Swords, dpf+:h:, RT, RT loops - Works if opponent is slightly out from you, but different bodies can effect the final hits fo swords before RT is thrown and late in combos they will flip out.
[/list]
I’m not too concerned with corner loops since the goal is usually to perpetuate midscreen ones and use sideswitching, so anything you can add to this would be greatly appreciated.