YesNoOk

Show content

This section allows you to browse the content for this member. Note that you can only see content for which you have sufficient viewing permissions.

**
Liero is Offline
Contact Liero:

Liero

Contributor

Messages by Liero

    

Re: Damage not calculating quickly enough

 November 22, 2013, 07:55:45 pm View in topic context
 Posted by Liero  in Damage not calculating quickly enough (Started by Liero November 22, 2013, 09:05:32 am
 Board: M.U.G.E.N Development Help

I do use that on occasion so that may actually be it.  I'll check once I'm back in front of it.

I imagine the best way to trigger hitdefs is through animelem, correct?
    

Damage not calculating quickly enough

 November 22, 2013, 09:05:32 am View in topic context
 Posted by Liero  in Damage not calculating quickly enough (Started by Liero November 22, 2013, 09:05:32 am
 Board: M.U.G.E.N Development Help

So I have a quick thing I'm trying to implement.  This is mostly for testing purposes to adjust and save some time before mass editing a bunch of characters.  Basically in -2 I have a float variable that changes based on which state you're in:

Spoiler, click to toggle visibilty

Now in the hitdef I'm doing this:

Spoiler, click to toggle visibilty

To clarify, fvar(13) is just for damage scaling, so that can be ignored.

The problem is this doesn't seem to be happening fast enough.  I'll do a normal standing jab and the damage will be doubled as expected, but if I do any special, the damage will ALSO be doubled on the first time I do the attack.  If I do it a second time, it will be the normal damage.  If I attempt to do another standing jab again, the damage output retains whatever modifier the last attack had until I repeat it or another attack in the group a second time.

At first I thought the variable simply wasn't changing fast enough, but using a move with super pause and debug mode, I have confirmed that the variable is assigning no later than the first frame.  Next I tried to re-order the operation of the damage calculation.  I tried this with the same results:

Spoiler, click to toggle visibilty

Any ideas as to why this seems to be happening too slowly?

EDIT:

On a whim I just tried forcing fvar(39) to my desired modifier in 2 separate states; one in the 200-699 range and one outside of it.  The multiplier sets and applies properly with both operations.  Perhaps it is the -2 going to slow despite the results of the super pause test.  Even still, I'm not sure why this is happening.


    

Re: Using frame data to balance your game +neat excel tool.

 August 18, 2012, 10:54:04 pm View in topic context
 Posted by Liero  in Using frame data to balance your game +neat excel tool. (Started by Liero August 11, 2012, 01:04:14 am
 Board: Development

snip

You know what?  I honestly thought you were just disagreeing with me for the sake of it.  Your last post convinced me that you actually know what you're talking about.  Granted I still disagree with you, but at least I don't feel like I wasted time trying to convince you.
    

Re: Using frame data to balance your game +neat excel tool.

 August 16, 2012, 07:24:39 am View in topic context
 Posted by Liero  in Using frame data to balance your game +neat excel tool. (Started by Liero August 11, 2012, 01:04:14 am
 Board: Development

snip

We have reached a point in this discussion where I don't think I can convince you of anything contradictory to your current stance.  However, I'll try to explain myself once more.

You say a character is not interesting because it can do a frame trap?  That simple thing by itself, no, but creating situations where you can give your character the tools to give options to their offense makes them interesting.  Frame traps create mixups.  Mixups create mind games.  Mind games create satisfaction when you can trick your opponent.  Characters who have more options are always more viable and more fun.  Without accurate frame data, you can't make a proper frame trap.  Sure you can guess and do trial and error, but as perfect as you think your execution may be, it's still not 100% and as such is error prone.  Here's an example:

A grappler's command throw is active on the 5th frame.
Jumping starts on the 4th frame
The grappler has an attack that on block is intended to give him enough advantage to make the grab easier.

Now here's where frame data comes in.

If you make that attack +2 on guard, the player cannot jump away in time, but if you make it +1, they can.  Right here you make an option and a mixup and as a result; a mind game.  The opponent knows they now cannot jump after blocking this move.  If you make that same attack +1 (or anything less than +2), the opponent can now jump out and the option is gone.  If you make that same attack with too much guard stun, the grappler now can't even grab the opponent (assuming you don't allow grabbing people while in hit stun).

This is a very simple example and there are many more things you can do with it such as giving the other players throw invincible backdashes and as a result giving the grappler a way to chase backdashes.  You're creating more interesting meta games by manipulating frame data effectively.  If you're just throwing numbers around, you lose these subtleties.

Of course there's more to games than just frame data.  I think you fundamentally misunderstand me.  I'm not implying that this is the end-all way to balance your game.  I never said that even once.  What I always said was that this is absolutely crucial.  Sure, work on your game mechanics first, but you cannot ignore this ESPECIALLY if you're making a full game.  You sent me private messages about some issues with the current build of my game.  Do you want to know how I'm resolving these issues to make the game better?  That's right; frame data.  All I can tell you is that the people involved in the closed testing for my upcoming patch are enjoying it much more because I have used frame data to create more options.

Design isn't about ticking boxes and going "how can a character deal with this or that"

Actually, that is PRECISELY what it is.  Have you played SF2HD Remix?  It's a re balanced version of Street Fighter 2 Turbo.  The lead designer of that game went through each individual character and analyzed situations where it was previously impossible for them to win and gave them a tool to deal with it.  I'm not talking dumb little annoying things, but I'm talking really really unbalanced problems.  For example;   a good Guile player in SF2T has an insane advantage over EHonda because EHonda has not safe way to approach.  HDRemix fixed this issue by making his jab sumo headbutt destroy projectiles.  The trade off is that he doesn't move very far, but he can still safely approach now.  Another incredibly simple example is what he did to Ryu.  Ryu had zero mind games and no way to trick anyone so the lead designer gave him a fake fireball.  That's it.  The fake fireball looks the same on startup but has much faster recovery.  That alone made him more viable. Neither of these are related to frame data specifically, but I'm using them as examples to explain why making proper design choices is essential to fighting games.

Poor design choices are what leads to games like SF3 3rd strike.  There's a very good reason why Chun Li, Ken, and Yun are pretty much the only characters you ever saw winning tournaments.  Most characters couldn't deal with Chun Li's back hard punch.  Yun was the only one that could.  Ken was just a decent character.  #rd strike wasn't a bad game, but it was pretty boring to watch tournaments.

I get what you're saying about "brave design choices", but you can make a great character while observing frame data.  Ensuring balance and creating options is not getting "trapped in the semantics of the genre".  You can build a beautiful house will still measuring things out and making sure you've got things built up to code. 

Finally, I have to say this again: this is not just my design approach.  I learned this approach from fighting game pros and people who have developed fighting games before.  If you seriously still disagree, that is fine and, ultimately, none of my business.  I have no energy left to try to convince you otherwise.


^I don't understand how his approach makes any of that stuff not possible? Frame data is a good starting point to add new mechanics like the ones you are talking about. And they too require their own frame data.

What he's saying here can be applied to SF, CVS2, JoJo, Guilty Gear, etc

You're just mentioning more options that the developer can add and they'd still need the same info. How much meter does x attack give to the burst meter, how many active ticks are the burst, etc...

The approach I am using can make everything he has mentioned.  This method only makes them less error prone ,less trial and error, but more accurate and more likely to work as the designer intended them to.

The biggest advantage of using frame data/accurate data (and I am not belittling the topic creator, if anything I support his points) is that the average mugen creators creates horrible frame data that completely ruins general gameplay and using a game as reference makes a tremendous difference.

Yeah and I've been guilty of this as well. 

Reference goes a long way, but if people aren't doing enough research, the accuracy is still going to suffer.

I wont say which one or which creator it was, but I recently downloaded a SF3 Ryu that was advertised as being 100% accurate.  I only checked his 3 punches, but the jab had 5 active frames too many, the strong was massively unsafe on hit, and the fierce took 3 frames longer to startup than it actually is supposed to.  I'm not going to tell anyone how to make anything.  All I can offer is my advice that you do a bit more research if you're using a game as a reference point.

    

Re: Using frame data to balance your game +neat excel tool.

 August 16, 2012, 12:42:49 am View in topic context
 Posted by Liero  in Using frame data to balance your game +neat excel tool. (Started by Liero August 11, 2012, 01:04:14 am
 Board: Development

I must not fully understand your ACT then. I thought it was the number of frames after the attack. I'm using your excel sheet mind you. I like this idea. I hope I didn't come off negative. I was merely trying to give a bit of feedback. Thank you for sharing this.

ACT refers to "active frames" or how long the attack is active for.  Basically you count how many ticks/frames the red collision is out.  Once the red collision is gone, your attack is no longer deemed to be active.

Look, it's not like you or any one else was particularly negative or overtly critical.  My disappointment comes from the people that view this sort of balance as unnecessary. I'm sure you know that MUGEN has a reputation for being unbalanced and generally silly in the fighting game community.  I was merely trying to offer my knowledge to everyone in the effort to help improve the overall quality of MUGEN games.  I guarantee you though, that I did not invent these techniques.  I learned how to do this from actual fighting game developers.

Glad you're getting use out of it though and glad that it's helping you.
    

Re: Using frame data to balance your game +neat excel tool.

 August 16, 2012, 12:21:22 am View in topic context
 Posted by Liero  in Using frame data to balance your game +neat excel tool. (Started by Liero August 11, 2012, 01:04:14 am
 Board: Development

I see what you're saying now.

Of course, you could do that on your own without this tool.  I used to do it that same way, but I wanted something easier to both calculate it and overview it without having to constantly fish around for data.  I keep track of this for every single character I make.  Yeah it's a lot of work and yeah it's time consuming.  I've stated a ton of reasons above as to why it's useful.  Do it if you want and don't if you don't.

For example Ryu's Light Punch.

DUR = 12
IMP = 4
ACT = 6

But I can get the same info by counting ticks and more.

frame 1 =3
frame 2 =3 attack frame
frame 3 =3
frame 4 =3


Be careful with this, but the way.  You have your active listed as 6, but your chart shows it as being 3 (unless there's an attack frame on frame 3 as well).

Admittedly, I'm a bit disappointed by how negatively people are responding to this, but I hope at least you get some use out of it.

    

Re: Using frame data to balance your game +neat excel tool.

 August 16, 2012, 12:00:07 am View in topic context
 Posted by Liero  in Using frame data to balance your game +neat excel tool. (Started by Liero August 11, 2012, 01:04:14 am
 Board: Development

    

Re: Using frame data to balance your game +neat excel tool.

 August 15, 2012, 11:49:01 pm View in topic context
 Posted by Liero  in Using frame data to balance your game +neat excel tool. (Started by Liero August 11, 2012, 01:04:14 am
 Board: Development

I like this for the most part. I think I'll use it to collect the data for my own character. 1 thing I don't see a big use for here though is the DUR IMP and ACT. For playing fighting games this is useful but if a person is going for more accuracy they'll be counting the ticks of every frame of animation. Unless this part is more for the players and not the creators. For creation purposes it's sort of redundant.

Actually, frame data is readily available for most popular fighting games out there.  Having played a great deal of them and having studied frame data on all of them, I've had little to no issue finding this information.

The irony being that DUR, IMP, and ACT are the 3 most important things on this sheet.  The funny thing is that it sounds like you're more concerned with creating something accurate than your own character.  If you were truly interested in doing that, you would absolutely need accurate frame data to do it.  This is mostly for you as a creator.  It's nice for some players, but ultimately this sheet is to help you make your game better.

I think you've kinda misunderstood me here; I don't have anything wrong with frame data being used. I'm trying to illustrate a principle.

What I'm trying to say is that you shouldn't be overengineering things. Players can do amazingly creative things with well-designed systems - and badly-designed ones too - but if you make it so that the only things that are possible are the things you want to be, you stifle that creativity and end up with a sterile game.

      Posted: August 15, 2012, 06:22:51 am
If you get things right at the fundamental system level, give your characters options, and fix things that are causing trouble, the specifics won't matter anywhere near as much. There are plenty of games around with weird quirks that haven't suffered from them (Jojo, CvS2, the more adventurous KoF games...).

(character limits, heh)

This is sort of problematic.  It's not over engineering; it's making sure there aren't problems/creating more combos/making your game more fun.  You also cite Jojo, CvS2, and KoF; all of which I guarantee you were developed in the exact same way.  If you're trying to convince me that Capcom and SNK don't bother with frame data, you are mistaken.

I like this for the most part. I think I'll use it to collect the data for my own character. 1 thing I don't see a big use for here though is the DUR IMP and ACT. For playing fighting games this is useful but if a person is going for more accuracy they'll be counting the ticks of every frame of animation. Unless this part is more for the players and not the creators. For creation purposes it's sort of redundant.

Gonna have to agree with this. While I'm not going to debate the usefulness of the spreadsheet, the IMP values are kind of redundant when it comes to creating stuff for mugen. The only real purpose I can see for this is if someone was making a character from scratch (Lets just say Reu's Dragonclaw or Seanaltly's Squall as arbitrary examples, as they have never appeared in a 2D fighting game outside of MUGEN) and wanted to calculate the frame data to give players an overall feel of character's moves without having to dig through his state files and calculate the framedata manually.

I could be mistaken, but I'm sure that the DUR & ACT values are used in calculating the frame dis/advantage (since no parameter for recovery frames exists, I'm assuming it's calculating this value by subtracting from the total duration).

Aside from compiling framedata for users, it can give the author an idea of how safe/unsafe a move is (disregarding hitvels and the like, of course) especially if the character itself is from scratch or the author otherwise doesn't have access to source game's frame data (above Dragonclaw/Squall example). Irregardless, I know this is something I'll be using.

This is actually useful if youre making any character regardless of whether it's been in a fighting game before or not.  Let's say you want to make a 100% accurate Ryu from SF3.  All you have to do is google search his frame data and hitboxes and you're half way there.  If you want to make a unique Ryu but still have him balanced, you can use this sheet to help you.

At any rate, I've come to the conclusion that many of you probably don't play that many fighters outside of Mugen; which ultimately is fine, but if you played competitively like I and some others do, you'd probably find a better use out of this tool.  If you'd rather go by feel or not look up data, you're not going to get any use out of this at all.
    

Re: Using frame data to balance your game +neat excel tool.

 August 14, 2012, 10:57:29 pm View in topic context
 Posted by Liero  in Using frame data to balance your game +neat excel tool. (Started by Liero August 11, 2012, 01:04:14 am
 Board: Development

Well,look; I wasn't implying that frame data is the only way to go when it comes to balance.  I even stated that frame data becomes irrelevant on knock down.  Of course you need some sort of way to handle juggling properly; that goes without saying, but flat-out ignoring this data is dangerous.  I absolutely guarantee that ignoring it will yield problems. 

Sure frame data can be slightly misleading, but only in the sense that a -6 move can still be safe based on push back, but knowing that your move is -6 on block helps you realize that you really should be throwing moves out after it's blocked regardless of your position.  You can also just trial and error this stuff if you want to find links and stuff like that but the effectiveness of this is purely limited by your level of execution.  If you can't do 1 frame links 100% of the time, then you need to use frame data to check your game.  I will agree with you, however, that more often then not, trial and error is the only way to balance velocities.

It's sort of hard for me to respond to your last paragraph without sounding rude, but I disagree.  Every successful fighting game ever made has based their balance mostly on frame data.  Sure, moving hitboxes around does a great deal towards balance, but it really mostly comes down to math and numbers.  If it didn't, then why would numbers even be used when coding these characters?  If it was all organic, then they wouldn't bother using so many variables.  This is 100% pure calculations whether you're measuring frame data, velocities, or the size of hitboxes.  If you're just throwing numbers around hoping for the best, I guarantee you that your character is either going to be broken or terrible.

If you still disagree with me about the necessities of frame data, pick any fighting game.  Seriously pick any one.  Google search for frame data for that game and you'll probably find it.  Now try doing the same for a good balanced game like SSF4 AE 2012.  If you understand the frame data, you'll be able to see why that game is so popular.  There are very few juggles in that game and it mostly relies on ground links for combos which makes frame data crucial for balance.  Yeah, maybe you wont need it as much for games with a lot of juggling, but you still need to pay attention to it for moves that either don't juggle or are supposed to be unsafe or punishable on block.

You also can't use execution limitations as a reason to not balance properly.  Sure a character may be so hard to use that their true power wont likely be realized by normal players, but there are many people out there that can do it all.  Take C Viper from SSF4.  She is hands down the most difficult character to play in that game, but she is easily one of the best in the right hands.  Viola from SCV is the same idea.  She is hard to use properly, but she's so good that there are talks of banning her from tournament play.  Look at her frame data and maybe you'll get why.  It doesn't matter how hard you make something.  If it is numerically possible  to do (frame data wise), a human can probably do it.  Seriously, the tightest timing you can possibly make a link is 1 frame.  Every single top SF4 player can do these 100% of the time.  I consider my self a slightly above average SF4 player and I can do it about 60% of the time.  This limit is not sufficient.

Again, I wasn't trying to sound rude, and apologies if I did.  My recommendation to the people that want to make something serious is to take the extra time to make sure your stuff adds up.  If you really don't care about making something balanced, then sure; don't bother.
    

Using frame data to balance your game +neat excel tool.

 August 11, 2012, 01:04:14 am View in topic context
 Posted by Liero  in Using frame data to balance your game +neat excel tool. (Started by Liero August 11, 2012, 01:04:14 am
 Board: Development

Ok so I saw another thread where someone was wondering how to calculate frame data in relation to hit/guardtime. If you're trying to make a balanced fighter or at least make an accurate clone of an existing character, this is essential.  For those who hate doing the math on their own, I made a nice little excel sheet to make this really easy to both calculate and keep track of all your character's data.  You can grab this sheet here: http://db.tt/12yTYTKa.  In this sheet, you'll find a template I used for my game (Rumble Pack : http://rumblepk.tumblr.com/) as well as the full character data for one of the characters so you can see what my thought process was or whatever. 

There's quite a bit to read, but I've been doing this for while and have studied this extensively.  Hopefully this will be helpful to you.

Chart Usage: Section 1 - How do you use this sheet? 
Spoiler, click to toggle visibilty

Chart usage: Section 2 - What's the other stuff for?
Spoiler, click to toggle visibilty

Why is frame data important?
Spoiler, click to toggle visibilty

Using frame data to create combos and frame traps
Spoiler, click to toggle visibilty

How to make/fix infinites or how to make people not want/want to play your game
Spoiler, click to toggle visibilty

Anything else?
Spoiler, click to toggle visibilty

Can you help with this?
Spoiler, click to toggle visibilty

Thanks for reading, and I hope this helps!

Sources/footnotes
Spoiler, click to toggle visibilty
    

Re: Changestate according to power

 June 30, 2012, 10:59:35 pm View in topic context
 Posted by Liero  in [SOLVED] Changestate according to power (Started by Liero June 30, 2012, 10:56:16 pm
 Board: M.U.G.E.N Development Help

Disregard.

Turns out I was missing a number next to "state"

*facepalm
    

[SOLVED] Changestate according to power

 June 30, 2012, 10:56:16 pm View in topic context
 Posted by Liero  in [SOLVED] Changestate according to power (Started by Liero June 30, 2012, 10:56:16 pm
 Board: M.U.G.E.N Development Help

This one is confusing me to no end.  I'm trying to have a dodge where, based on how much power you have, you have more or less recover frames.  Here's the snip of code I'm using:

Code:
[state, recover]
type = changestate
trigger1 = power = [1750,9999]
trigger1 = time = 20
trigger2 = power = [1500,1749]
trigger2 = time = 22
trigger3 = power = [1250,1499]
trigger3 = time = 24
trigger4 = power = [1000,1249]
trigger4 = time = 26
trigger5 = power = [750, 999]
trigger5 = time = 28
trigger6 = power = [500, 699]
trigger6 = time = 30
trigger7 = power = [250, 499]
trigger7 = time = 32
trigger8 = power = [0, 249]
trigger8 = time = 34
value = 0
ctrl = 1

For some weird reason, this does not do anything.  The character simply loops over and over again in the state and that's that.  I've tried omitting references to power and it works just fine, but no matter what I seem to do, changing states directly based upon the player's power does not work for me.

Is there some other way to get this to work?

PS

I've also tried assigning power to a variable in the player's -2 state and triggering it that way, but that didn't work either.
    

Re: Rumble Pack!

 February 03, 2012, 01:17:21 am View in topic context
 Posted by Liero  in Rumble Pack! (Started by Liero June 20, 2011, 10:51:58 pm
 Board: Your Releases, Mugen 1.0 +

    

Re: Canrecover issue

 January 26, 2012, 08:12:35 am View in topic context
 Posted by Liero  in [SOLVED] Canrecover issue (Started by Liero January 24, 2012, 10:09:29 pm
 Board: M.U.G.E.N Development Help

I think what you want is overriding these in common1.cns?

Quote
[State 5050, 4] ;Recover near ground
type = ChangeState
triggerall = Vel Y > 0
triggerall = Pos Y >= Const(movement.air.gethit.groundrecover.ground.threshol d)
triggerall = alive
triggerall = CanRecover
trigger1 = Command = "recovery"
value = 5200 ;Air get-hit (fall recovery on ground)

[State 5050, 5]; Recover in mid air
type = ChangeState
triggerall = Vel Y > Const(movement.air.gethit.airrecover.threshold)
triggerall = alive
triggerall = CanRecover
trigger1 = Command = "recovery"
value = 5210 ;Air get-hit (fall recovery in air)
They are in state 5050, which is the (only) standard falling state.
If you also want to recover from custom falling states, you'll need the code in state -2, you'll need to use selfstate, and you'll need to code it like this

[State -2]
Type=Selfstate
Triggerall=Movetype=H
Triggerall=hitfall
Triggerall=canrecover
Trigger1=command="recovery"
value=your recovery state number

Yep, this one worked.  Thanks!
    

Re: Canrecover issue

 January 25, 2012, 01:31:56 am View in topic context
 Posted by Liero  in [SOLVED] Canrecover issue (Started by Liero January 24, 2012, 10:09:29 pm
 Board: M.U.G.E.N Development Help

Yessir.  It's one of those fly-across-the-screen-until-you-hit-the-wall states
    

[SOLVED] Canrecover issue

 January 24, 2012, 10:09:29 pm View in topic context
 Posted by Liero  in [SOLVED] Canrecover issue (Started by Liero January 24, 2012, 10:09:29 pm
 Board: M.U.G.E.N Development Help

So I'm trying to use "canrecover" in a state, but as you know, it can only be used in a falling state otherwise it doesn't do anything.  How do I define a state as a falling state?  Or rather, how can I use canrecover (or some sort of equivalent) outside a falling state?

Currently, what I have sort of works.  I am using this:

Code:
[State 5100, tech]
type = ChangeState
triggerall = time < 12
triggerall = gethitvar(fall.recover) = 1;canrecover = 1
trigger1 = command="a" || command="b" || command="c" || command="x" || command="y" || command="z"
value = 8065

The problem is that it only works if the attacker is still in his attack animation.  Once it has finished, it no longer works.
    

Re: Rumble Pack!

 January 23, 2012, 11:56:17 am View in topic context
 Posted by Liero  in Rumble Pack! (Started by Liero June 20, 2011, 10:51:58 pm
 Board: Your Releases, Mugen 1.0 +

    

Re: Rumble Pack!

 January 23, 2012, 02:28:59 am View in topic context
 Posted by Liero  in Rumble Pack! (Started by Liero June 20, 2011, 10:51:58 pm
 Board: Your Releases, Mugen 1.0 +

    

Re: Rumble Pack!

 January 21, 2012, 12:42:14 pm View in topic context
 Posted by Liero  in Rumble Pack! (Started by Liero June 20, 2011, 10:51:58 pm
 Board: Your Releases, Mugen 1.0 +

    

Re: Rumble Pack!

 January 21, 2012, 02:40:43 am View in topic context
 Posted by Liero  in Rumble Pack! (Started by Liero June 20, 2011, 10:51:58 pm
 Board: Your Releases, Mugen 1.0 +

-It's not a training mode only combo. That's 3 meter and 3 meter is the maximum.
-I think Nadia's Standing C is buggin out as well.
-Binary Nadia's EX QCF move does ridiculous damage. More than her super.

Oh yeah you're right.  That style actually used to be her Hex style (maxes at 2) so I got confused.

I'll check that move too

Wow, that is a big problem.  I'm putting out an emergency patch tonight anyways, so I'll include this on it.  Thanks for the catch.

EDIT:

So for the Charlotte thing; turns out that that super was accidentally resetting the damage scaling.  I've also made it a bit more difficult to combo out of that move.