• As part of the relaunch of Skullheart, ALL previous threads have been archived. You can find them at the bottom of the forum in the Archives (2021) section. The archives are locked, so please use the new forum sections to create new discussion threads.

How fast are your resets? Find out here!

guitalex2007

Well-Known Member
Joined
Sep 8, 2013
Messages
978
Reaction score
1,112
Points
93
Age
41
Steam
guitalex2007
PSN
guitalex2007
XBL
guitalex2010
I'll keep this as short as possible.

So you've always wondered how fast your reset is?

OMG MUST RESET BUT DON'T KNOW HOW LONG THE RESET IS ARGH.

There is a way to find out EXACTLY how many frames there are between the hits in a reset! And for those of you who haven't figured it out, once you read this short explanation you'll facepalm because of how damn simple it is!

The method involves doing the combo in training mode (obviously) with attack data on. Advanced attack data is optional.

Drama/Undizzy will show next to combo stage as xx/240, where 240 is the max. From 240, it takes 40f to decay, so you could say drama decays at a rate of 6 drama per frame.

(Not click yet? OK, read on)

Drama will build during a combo, and will not go up in stage 1/2. So the drama will go up until the opponent goes to neutral, and will not increase in stage 1/2.

(Not yet?)

So resets will stop the drama/undizzy from going further down after a reset... so all you have to do is calculate or write down how much drama/undizzy the combo has during the last hit, reset and write down the drama/undizzy and subtract both and divide by the decay rate of 6 undizzy per frame.

HERE'S MATH.
A Fukua combo.

c.LK c.MK s.HP (undizzy starts here) j.HK s.HP j.HP j.HK s.MK xx qcb+throw

The s.MK into a throw is a reset. How tight is it?

HOW WILL I EVER KNOW?

When the s.MK hit, undizzy was 140. During the throw, I checked and it said 101. So... 140-101=39 undizzy decayed. So that is 6.5f (39/6 = 6.5) so if you want to round up, 7f reset. In seconds, divide by 60. Don't forget to account for frameskip.

BOOM.

Yeah, that was the sound of some of you facepalming for not noticing this before.

WHAT? MATH!? SCREW MATH!

Go here then. Thank @Vadsamoht for that one.

ENJOY.
 
Last edited:
There. Better?
 
I read this, and my lazy nature got the better of me.
I read it a second time and everything clicked.

Maybe I should draw some diagrams or something.
 
(Unidizzy before the reset - Undizzy after the reset)/6 = frames between the reset not counting frameskip

That would be the TL;DR version.
 
It'd be pretty easy to make a calculator for this. What's the current frameskip value?
 
Every 7 frames the next one is skipped. Problem is you don't know which frame is skipped but you can still approximate plus one or two.

In a 1-7f window there's a chance one of those is skipped. In an 8f window one of them is definitely skipped. But in a 9-15f window one or two will be skipped for sure. For a 16f window exactly two frames are skipped. Then at 17-23f two or three are skipped. Don't think we need any more than that.

So if the result is like explained before 6.5f which I rounded to 7f, the reset is 6 or 7f perceived due to frameskip. A 10f reset is really 8 to 10f perceived depending on which frames skip.
 
Last edited:
Ok, simple calculator is up here. It's not fancy and doesn't check for bad input, but should work otherwise.

EDIT: If someone could test it out and let me know that it actually works, that'd be great.
 
Last edited:
Ok, simple calculator is up here. It's not fancy and doesn't check for bad input, but should work otherwise.

EDIT: If someone could test it out and let me know that it actually works, that'd be great.
Seems to do the math right. EVEN EASIER!

WHY DO MATH EVER?

(Other than for programming things to do math for you, that is...)

Doesn't do the math for up to 240 but why would anyone want a 60f reset is beyond me.

Also, would be nice to see the reset in ms since that way we can know if the resets are reactable. The math is:

Perceived frames after frameskip / 60 * 1000 = time in ms

(Not that this helps any against mashing but there's something to compare it to)
 
Last edited:
Ok, I think I've fixed the problem with large resets and I've added a display for the duration of the reset in ms.

EDIT: lolninjas
 
Ok, I think I've fixed the problem with large resets and I've added a display for the duration of the reset in ms.

EDIT: lolninjas
OK now I get that at 3f (199-182) says 3 actual frames, but it says with frameskip is 1 or 2. Should be 2 or 3.
 
Ok, I think the equation should be pretty watertight now.
 
I think there might be a slight problem with your algorithm, Guitalex. I've got a reset that has no drama decay (115 before, 115 after), which with the current equation would mean that the reset takes 0 frames and is thus part of the same combo. Are you sure that the Drama decay doesn't update after each frame, meaning that every calculated value so far is actually 1 frame less that the actual length of the reset?
 
I think there might be a slight problem with your algorithm, Guitalex. I've got a reset that has no drama decay (115 before, 115 after), which with the current equation would mean that the reset takes 0 frames and is thus part of the same combo. Are you sure that the Drama decay doesn't update after each frame, meaning that every calculated value so far is actually 1 frame less that the actual length of the reset?
0f resets are possible, but I wonder if drama stays decaying immediately or not.

@Mike_Z
 
It looks like it waits 2f after you recover to start decaying at all.
Is this...intentional?
 
So instead of being (Drama before - Drama after)/6 it's actually [(Drama before - Drama after)/6] + 2? If that's it then it's pretty simple to update.
 
Ok, Calculator updated.
 
Nerds
You're the world champion at Skullgirls.

Pot, meet kettle.
 
  • Like
Reactions: Midiman
I did an .exe version of this.
Mainly to learn more stuff about Delphi stuff, like how to use OnKeyPress events and this kind of stuff.
This link is a direct link to the download. It's inside a .rar archive.