GGST/Miscellaneous

Finding content on Twitter

With the ease of posting videos and discussion to Twitter, it can be helpful to know how to track this information down. Each character in the game has a specific hashtag which people should aim to use when posting tweets for ease of searching. The current list of these hashtags is as follows:

Sol Badguy	#GGST_SO
Ky Kiske	#GGST_KY
May	#GGST_MA
Axl Low	#GGST_AX
Chipp Zanuff	#GGST_CH
Potemkin	#GGST_PO
Faust	#GGST_FA
Millia Rage	#GGST_MI
Zato-1	#GGST_ZA
Ramlethal Valentine	#GGST_RA
Leo	#GGST_LE
Nagoriyuki	#GGST_NA
Giovanna	#GGST_GI
Anji Mito	#GGST_AN
I-No	#GGST_IN
Goldlewis Dickinson	#GGST_GO
Jack-O'	#GGST_JC
Happy Chaos	#GGST_HA
Baiken	#GGST_BA
Testament	#GGST_TE
Bridget	#GGST_BR

Input Shortcuts

Certain specials and Overdrives allow you skip directional inputs for their commands and still execute the attack. Below are the known valid input shortcuts for their usual directional commands (additional directional inputs within the given shortcuts will also work, as long as the minimum inputs as shown in the shortcut are met, e.g. 62146 is valid because 6246 is valid).

Sol performing Heavy Mob Cemetery (214214H) with the input shortcut 2421H

Command	Shortcuts
41236 (63214)	4236 4126 4136
236236 (214214)	2626 2623 2636
632146	6246 63146
632147896 (360 motion)	6248 4268

Low-End Considerations

Strive is a graphically-extensive game that will greatly bar out the lower end of older computers, needing much more power as opposed to Xrd just to run minimal settings (seeing as the game usually takes up around 4GB of RAM on its own). There are a few things you can try to alleviate any issues with stability in running the game, but there is only so much you can do before you're out of luck.

Settings

Quality 0, looking like blurred pixels. Sometimes, they won't be blurred and actually pixelated, but you don't want to play like this.

Firstly, be sure to set your resolution to lower than the default 1920x1080, such as 1280x720.

As usual, set your graphics setting all to the lowest values/OFF minus the Scaling Resolution and Effect Quality. Lowering Scaling Resolution will severely lower the pixel density of the in-game 3D graphics on your screen to speed things up along and use less power, but obviously this will adversely affect how you view gameplay - while it's still technically playable at 0, you'd have to question your decisions at that point. The minimum you might want to handle is 50, depending on your tastes and tolerance, with 75 being a good mix of seeing what's happening and performance. You'll have to play around with these values and judge it for yourself with screenshots. Note that this does not affect UI elements like text, gauges, buttons, so on.

Lastly, at the time of writing, there seems to be an oversight where not setting your Effect Quality to Best gives a very strange and ugly bloom/lighting effect on characters. However, you do not have to keep the setting at Best! At any point of the game, all you need to do is press Alt and Enter on your keyboard, which will toggle your fullscreen. Do just that and the bloom issues should no longer be present for the game's current session.

Scalability.ini

Thanks to this article's observation, another step can be made to slightly increase performance of your game. Make sure your game is closed, then navigate to Windows Explorer and type in %localappdata%. Hit enter and it should bring you to a folder called Local. In there, find GGST -> Saved -> Config and lastly open the Scalability.ini file that's in there. Copy and paste the following contents into there, then save the file.

[AntiAliasingQuality@0]

r.MSAA.CompositingSampleCount=1

[ViewDistanceQuality@0]

r.SkeletalMeshLODBias=5

r.ViewDistanceScale=0.0

[ShadowQuality@0]

r.ShadowQuality=0

r.Shadow.CSM.MaxCascades=0

r.Shadow.MaxResolution=0

r.Shadow.RadiusThreshold=0.0

r.Shadow.DistanceScale=0.0

r.AllowLandscapeShadows=0

[PostProcessQuality@0]

r.MotionBlurQuality=0

r.AmbientOcclusionMipLevelFactor=0

r.AmbientOcclusionRadiusScale=0

r.RenderTargetPoolMin=200

r.BloomQuality=0

r.Upscale.Quality=0

r.Filter.SizeScale=0

r.Tonemapper.Quality=0

[TextureQuality@0]

r.Streaming.MipBias=3

r.MaxAnisotropy=0

r.Streaming.PoolSize=800

[FoliageQuality@0]

r.ParticleLightQuality=0

Next, you probably don't want the game to override these settings if you're satisfied, so right-click the ini, go to Properties, and set the Read-Only flag to checked and confirm it.

Stages

Unfortunately, not all stages in Strive are built the same, with some being more hardware intensive than others. This can be very detrimental when playing online with strangers especially if you're on Player 2 side, as having performance issues such as frame drops or lag spikes will leading to high rollback frames or worse the game slowing down. Observations on which stages are more intensive than others can be compared here, so you can use this to tell your friends which stages you can play on better if you're the low-end one out. However, remember that PC hardware is not unified so one stage might not give your PC any problems but it might give someone else's PC problems - these are all relative at best, but stages listed as A-OK should be the least intensive. So if you're experience some form of netplay lag on a specific stage try to check with an FPS counter like the one in the Steam overlay if your PC is running at a completely stable 60 fps and if possible ask the opponent to do the same.

Stage	Lagginess
Castle of the Silver Feet	Little Lag
Lars Canyon	A-OK
Seventh Heaven District	A-OK
12-Dimensional Flux Observatory	A-OK
Ajatar Hunting Ground	Very Laggy
"A Happy Family"	A-OK
L'oro Di Illyria	Some Lag
Council of Three	A-OK
Jellyfish Skyarium	Very Laggy
Original Times Square	Some Lag
Lap of the Kami	Very Laggy
White House Reborn	Some Lag

Note: White House Reborn is generally banned in tournament play, because the rubble that spawns after a Wall Break stays on screen for too long, obscuring post-Wall Break interactions.

Using Mods

The Lite version. Part of the stage is still intact, for average computers.

The original edits. The stage is destroyed and should no longer pose a threat to your potato.

Thankfully, modders have jumped into action to customize things like audio, models, and textures, but fortunately this also gives a possibility of fixing the offending stages for low-end setups. Using the following mods created by DraftFGC, much of the resource-hogging models on stages will be removed and not be rendered, which is a better visual alternative to completely removing all of the stage's characteristics.

This page has the most recent and most recommended edits, with per-stage selections that remove much of offending objects within stages without sacrificing the overall appearance like the other mods do. You should choose from this page if your slowdowns are mild. Alternative mirror here.

The first set of removals are here, which are collective and also remove skies and Character Selection Screen assets. Individual and manual downloads (without those extreme edits) can be found in this MEGA folder. For if Lite really doesn't fix it - this should do the job.

Another mod to make Ajatar Hunting Ground into a solid color exists (for green-screening), but you can only load one of either version of the mod.

After downloading the stages you've chosen, to manually install these mods, follow these steps:

Navigate to C:\Program Files (x86)\Steam\steamapps\common\GUILTY GEAR STRIVE\RED\Content\Paks\ (whichever drive/directory your game is saved in)
In the Paks folder, make another folder called ~mods
Extract and drop your downloaded contents into that folder and test in-game (in other words, you should have Paks/~mods/*.pak and *.sig).

You should no longer have to worry about fighting strangers or asking people to swap stages with these mods. Unfortunately, though, if your game is still having issues running even after all these steps, then there likely isn't any more you can do aside from upgrading your setup.

Training Mode Shortcuts

Direction + Reset: Left + Reset to reset both characters to the left corner. (P1 in the corner); Right + Reset to reset both characters to the right corner. (P2 in the corner); Down + Reset to reset both characters to round start position.; Up + Reset to reset both characters to the previous starting position with their player sides switched.

Play + Record = Reset: Pressing both Play and Record at the same time is the same as pressing the Reset button.

Technical Calculations

These are not required to utilise game mechanics, but these sections further elaborate on the mathematical calcuations for various system mechanics, which are otherwise too complex for their respective sections.

Base R.I.S.C. Gain

Base R.I.S.C Gain is tied to Attack Level, using the formula:

⌊(5 + 5 × Attack Level) ÷ Number of hits⌋ × 100

While values are normally rounded down before the ×100 multiplier, they will instead round up to 1 if the result is lower than 1.
Projectiles ignore Attack Level for their base R.I.S.C. Gain, acting like Level 0 attacks regardless of their actual level.
A fully charged Dust Attack functions like a Level 4 for the opponent on block despite being Level 2, including R.I.S.C. Gain.
R.I.S.C. Gain is reduced by 25% if the blocked attack is a low or an overhead.

R.I.S.C. Damage Reduction

When striking an opponent with a negative R.I.S.C. Level, the output damage is reduced. This is colloquially known as "combo scaling".

As a first approximation, take the table below, and see the "percentage" column - that is how much the damage is scaled for respective R.I.S.C. Level.

R.I.S.C. Level	Factor	Percentage
-100	256	100.00%
-1700	200	78.13%
-3300	152	59.38%
-4900	112	43.75%
-6500	80	31.25%
-8100	48	18.75%
-9700	32	12.50%
-11300	16	6.25%
-12900	8	3.13%

The exact math is as follows:

Find the two closest R.I.S.C. Level values in the left column of the table above, look up the corresponding values in the factor column.
Do the linear interpolation between the values obtained; round down to nearest integer.
Multiply the base attack damage by the number obtained, then divide by 256 and round down.
The result would be the damage taken by your opponent.

Example, in prose
Let's say we are landing a 30 damage attack on an opponent with a R.I.S.C Level of -4000. The two closest R.I.S.C. Level values in the table above would be -3300 (factor of 152) and -4900 (factor of 112). For simplification purposes, the negatives can be omitted. We then do the linear interpolation of the two (i.e. "finding the average of the values, accounting for the proximity", roughly) as follows: factor_a + (factor_b - factor_a) * (risc - risc_a) / (risc_b - risc_a) = 152 + (112 - 152) * (4000 - 3300) / (4900 - 3300) = 152 + (-40) * 700 / 1600 = 152 - 17.5 = 134.5 Round it down, getting 134; and get the final answer as `30 * 134 / 256 = 15.7 = (rounding down once more) = 15` damage.

Pseudocode

 fn scale_attack(attack_damage: int, risc: int)
   if(risc >= 0) {return attack_damage;}  // positive risc does not affect damage at all
   if(risc >= -12800) {risc = -12800} // can't go lower than this
   risc = abs(risc); // flip the sign so it's easier to work with
   // table from above, with sign flipped
   let risc_factor_table = {100: 256, 1700: 200, 3300: 152, 4900: 112, 6500: 80, 8100: 48, 9700: 32, 11300: 16, 12900: 8}; 
   let risc_ordinal = floor((risc - 100) / 1600);
   
   let risc_lower = risc_ordinal * 1600 + 100;
   let risc_upper = (risc_ordinal + 1) * 1600 + 100;
   let factor_lower = risc_factor_table[risc_lower];
   let factor_upper = risc_factor_table[risc_upper];
   // linear interpolation
   let factor_interpolated = factor_lower + floor((factor_upper - factor_lower) * (risc - risc_lower) / (risc_upper - risc_lower));
   
   let out_damage = floor(attack_damage * factor_interpolated / 256);
   return out_damage

Dash Acceleration

For each frame a dash is held (excluding Leo, Giovanna, and I-No's unique dashes), the dash speed will be:

vₙ₊₁ = ⌊vₙ + Acceleration - (vₙ/Friction)⌋

Where:

vₙ is the current speed
vₙ₊₁ is the speed next frame
Acceleration is the character's base acceleration
Friction is the character's base friction value
v cannot exceed 38.5
The value in ⌊ ⌋ is rounded down to the nearest 0.001

Initial dash speed is the value of v when n=0, therefore frame 1 will always be slightly faster. For the sake of calculation, friction is inverse, a higher value maintains speed, while a lower value loses acceleration quicker.

Due to rounding and the mechanics of friction there is no simple way to estimate speed from an arbitrary frame, without calculating the entire series.

Additional Data Charts

6P Values

6P Values

chara	startup	active	recovery	onBlock	onHit	invuln
chara	startup	active	recovery	onBlock	onHit	invuln
A.B.A	11	6	20	-12	-9	1-2 Upper Body/3-16 Above Knees
Anji Mito	10	6	25	-17	KD +19	1-3 Upper Body 4-15 Above Knees
Asuka R	11	6	17	-9	KD	1-3 Upper Body 4-16 Above Knees
Axl Low	12	5	22	-13	+2	1-3 Upper Body 4-16 Above Knees
Baiken	9	5	19	-10	-7	1-2 Upper Body 3-13 Above Knees
Bedman	11	6	24	-16	KD +19	1-2 Upper Body 3-16 Above Knees
Bridget	10	6	20	-12	KD +23	1-3 Upper Body 4-17 Above Knee
Chipp Zanuff	9	5	25	-16	-13	1-2 Upper Body 3-13 Above Knee
Elphelt Valentine	9	4	20	-10	KD +25	1-2 Upper Body 3-12 Above Knee
Faust	9	3	32	-21	KD +21	1-3 Upper Body 4-11 Above Knees
Giovanna	12	8	15	-9	KD +22	1-3 Upper Body 4-19 Above Knees
Goldlewis Dickinson	12	6	27	-19	-16	1-2 Upper Body 3-17 Above Knees
Happy Chaos	9	5	25	-16	-13	1-3 Upper Body 4-13 Above Knees
I-No	9	5	18	-9	-6	1-2 Upper Body 3-13 Above Knees
Jack-O	10	6	20	-12	KD +21	1-3 Upper Body 4-12 Above Knees
Johnny	11	6	17	-9	KD	1-3 Upper Body 4-16 Above Knee
Ky Kiske	9	5	17	-8	-5	1-3 Upper Body 4-13 Above Knees
Leo Whitefang	10	5	17	-5	+7	1-2 Upper Body 3-14 Above Knees
May	12	6	18	-7	KD +25	1-3 Upper Body 4-17 Above Knees
Millia Rage	9	9	16	-11	-8	1-2 Upper Body 3-11 Above Knee
Nagoriyuki	12	5	15	-6	-3	1-3 Upper Body 4-17 Above Knee
Potemkin	11	7	25	-15	KD +17	1-3 Upper Body 4-17 Above Knees
Ramlethal Valentine	9	5	18	-9	-6	1-3 Upper Body 4-13 Above Knees
Sin Kiske	11	6	20	-12	-9	1-3 Upper Body 4-16 Above Knee
Sol Badguy	9	5	20	-11	-8	1-2 Upper Body 3-13 Above Knees
Testament	11	6	15	-7	KD +28	1-3 Upper Body 4-17 Above knees
Zato-1	12	8	12	-6	KD +23	1-3 Upper Body 4-15 Above Knees

Reversal Values

Reversal Values

chara	input	startup	active	recovery	onBlock	onHit	invuln
chara	input	startup	active	recovery	onBlock	onHit	invuln
Anji Mito	632146S	1	30	10			Guard 1-30
Axl Low	632146P Attack	8+1+0	1	11	0	+140	8-9F Full
Axl Low	j.632146P Attack	8+1+0	1	11	0	+140	8-9F Full
Giovanna	632146H	5+4	4(4)4(4)4(4)4(4)7	49	-37	HKD +18	1-10F Full
Goldlewis Dickinson	632146P	6+(135 Flash)+4	5	46	-33	HKD +62	1-14F Full
Goldlewis Dickinson	720P	11+(191 Flash)+4	5	46	-33		1-19F Full
Goldlewis Dickinson	1080P	16+(230 Flash)+4	5	46	-33		1-24F Full
Ky Kiske	623S	11	4	43	-33 [-28]	KD +22 [KD +34]	1-14f Strike
Ky Kiske	632146H	8+1	3×4(20)2	99	-82	HKD +20	1-10F Full
Ky Kiske	DI 632146H	8+1	2×8(20)2×3	99	-79	HKD +26	1-15F Full
Potemkin	236236S	12+0	15	16	N/A	HKD +65	1-12F Full
Ramlethal Valentine	236236S	7+2	6	51	-38 [-38] {-43}	HKD +28	1-14F Full
Sin Kiske	632146H	11+5	12	25+32 After Landing	-55	HKD +20	1-18F Full