Good read! I was just wondering, if you had all this frame data and just knowledge you’ve accumulated from experience about what links and what doesn’t, wouldn’t it be possible to take all that information and create a program that would spit out lists of “possible” combos? Not saying that they would all work but you would be able to generate “possible” combos on a mass scale as opposed to just thinking about it. Might even be able to generate stuff that you wouldn’t have thought of. Again I’m not a combo maker so I don’t know what it’s like to create a combo that “works” and how it relates to hard game data and situational circumstances. But just a thought.
damn i seriously cant wait to see this!!
Hey young tro, did he document how long it took to complete each combo?
If I were y’all I’d add in a footnote at the end of each combo that states how long it took.
Or is that portion of the explanations already complete?
P.S. I didn’t see you mention much on positioning of the dummy. There’s also problems that arrive when something is ONE frame off which results in a dummy being too far or too close to the combo-er. If that happens really far into the combo, you could have a lot of backtracking to do to find where that ONE frame difference was!
I’ve come to learn how incredible it is (what Joo has actually done) so I feel like it would be wrong not to include these things for the average gamer to understand.
Sort of. Once you have the frame data you need to calculate for hitstop and all these other variables that would make the combo not work. It’s hard enough to come up with combos when you know the data.
There are also the combos that mathematically would work, but in order for them to work, you’ll have to do magic with positioning. The Cyclops vs Roll combo in Meikyousisui11 is a REALLY good example of position manipulation. Also, the Captain America FSD from DVDPreview.avi–the SJ.8HK was done early and far away so that he could move forward and hit with it late. He also did the Sj.8HK off the MK which moves the character downward, so if Captain America did the whole air combo soon after jumping, he wouldn’t have fast velocity that you get when you reach your apex and fall to move him down to ALLOW his 8HK to connect once he edged left. It’s really really remarkable when you see it explained.
Yeah, that’s what I was planning on doing with some of the combos. He told me the Cyclops56 (off-screen) combo took him a whole month or so during the summer of 2003.
The FSD combos (most likely the Scroll-FSD ones) took a long time because there are 4 important variables that all have to work out correctly in order for the combo to work.
*FS-Dash speed and duration & attack that is being used to follow up.
*Dummy Positioning before the combo
*Dummy positioning after the combo
*Main character’s position before the combo.
If any of those are off by one frame the Scroll-FSD will not work. Of course, you also have to know how long the dummy is going to be on the ground for. That can be found by mining frame data. So the way joo does it the most logical way when you think about it. Try each and every single possible variable. This could mean over one thousand numbers to try…on a combo that doesn’t work right 100% of the time. So lets say 2500 tries go in for a combo that requires 1000 variables to be tested. Joo has about 30 excel sheets with Combos and Data. Almost each excel sheet has about 15 sheets that are filled out with all kinds of data for each character. So 56 rows and 56 columns per sheet. Sometimes much more.
Right, the positioning in marvel is by far the biggest and hardest problem to overcome. But when you do, you get combos or situations that are very special. Some of his combos use double 100% technique to insure that both characters will start moving perfectly on exactly the right frame. For instance, the Ryu27Hit Scroll FSD from DVD Preview used the double 100% technique because the positioning required in order for the follow up had to be perfect. Each character had to move exactly a certain amount of pixels in order for the OTG to work.
So yeah, there are 130 combos in this DVD. Of course, not all of them took 3 months to make, but still.
so fucking DEEP
good read
Thanks for the info, please keep us updated.
Any update as to how the project is coming along?
Just saw the announcement of the end of January target release. Looking forward to it.
If its free…
Donation fund to Joo, now…
this is the updated thread with videos and stuff http://forums.shoryuken.com/showthread.php?t=157266&page=6