Experienced Forum User, Published Author, Player
(223)
Joined: 2/3/2016
Posts: 6
My fault, haha. I would be very interested in seeing a TAS that would go that high. The biggest issue (and the reason it would take a good, long time) is that points do suffer from decay. Points do not count the same the second, third, or fourth try and so on in the same combo or trick attack run. After a certain point, as well, a trick would only add a .5 to the multiplier.
Now, there are spots where we could be in the air forever, but the game actually realizes this after a little bit and stops any tricks from appearing in the trick string.
Thanks for the insight, Fog! :)
Experienced Forum User, Published Author, Player
(223)
Joined: 2/3/2016
Posts: 6
CATBAG stands for Create-A-Trick Broken-Ass-Game and it takes advantage of a flaw in the game's trick string you see in combos. The game doesn't check for if you actually did the correct trick, just for if the trick name itself is in the combo. So, as you saw throughout the video, you would have a Nut Buster count as the Nut Buster in Manhattan, a random Nollie trick would count as a real Nollie in Vancouver, and so on.
This was originally discovered by PokemonAce101 in THUG2 as a means to get around a frustrating goal. When it proved to kinda work, another runner, SubZeroReptile, took it into THUG1 and found it had an incredibly high potential in the game. And, from there, we worked together on finding methods that worked properly.
For starters: This method would allow us "copy" legit trick names and create new Created Tricks. This immediately meant we could replace the McTwist in our Speical Trick slots with something better, and instead have a Created Trick called McTwist. The game would think we did it when, obviously, we don't.
After that, we learned how to manipulate our trick sets into having multiples of the same trick. By choosing another unrelated Created Trick and overwriting it with the real one we wanted, the game wouldn't initiate a check to make sure we didn't have multiples of the same trick. This is why you see Nut Buster A LOT in this run, and why some goals *might* be a bit slower since we are limited to our special tricks as the only way to get points.
Third, this also meant we basically eliminated RNG in the run. In New Jersey 1, Tampa, and New Jersey 2, we had some big RNG segments that could either give us a really good seed, or a really bad one. Now, since literally every trick is called Nut Buster on both grab and flip (save for Madonna/Judo), whenever the game calls out multiple Nut Busters on the right side, one Nut Buster does it for all of them. This feature is intended by the game, but we just took it a step further.
Hope that explains that part!
As for executing tricks crazy fast, we are allowed to have Created Tricks have a 0.4 second long animation. No method has been found to reduce that, unfortunately, but it still looks really fast.
For the huge combo, if we did CATBAG, it would take hours, at the minimum. The Tony Hawk community does have people who go for points or huge lines, and the biggest I've ever seen 6 billion in THPS4 (and that took 20 or so minutes to accomplish?). So, is there a limit? I have no clue, actually, haha.
Experienced Forum User, Published Author, Player
(223)
Joined: 2/3/2016
Posts: 6
As I mentioned, NJ is less optimized then other areas since I was still getting the hang of TASing at the time. But, from watching it a half dozen times, that is the only area I would say is sub-par. Everything else was pretty good overall, with the exception of maybe Manhattan's call-out section and that is because I couldn't find a good seed.
The seed I got was the best I could go with at the time, because it changes all of the time (every ten frames or so) and it changes mid-mission too. I spent a couple of hours trying to map out a good call-out section, but I couldn't figure it out. I could've buttslapped during it, though, to get more tricks in but, by the time I realized it, I was already too far in.
Buttslapping is an one frame on, one frame off trick. After you do an ollie off of an edge and do a trick, you have a good twenty or so frames to actually do a buttslap. Normally, a human can only do three to four max. But, obviously with a TAS, you can do up to ten or so. All it is is hitting Ollie on one frame, letting go on the next, ollie again, off next, and so on until I got what I wanted.
As for the tricks counting, it is more a reference to my RTA runs because I do them all of the time and never have many count. I really just did it for fun because I hardly had anything else to do at the time.
Experienced Forum User, Published Author, Player
(223)
Joined: 2/3/2016
Posts: 6
Fog wrote:
MUGG wrote:
Oh this will be interesting to watch.
I'd argue that I did some of the New Jersey stuff slightly faster (car driving, combo clearing), but you really surprised me with some strats I hadn't considered. Will watch the rest later
Edit:
I watched till 12:00 so far. I often see you gain super speed but then you lose it... Is there a cap where the game reduces your speed if you go too fast? Something like that also happens in f-zero GX.
wtf @ jumping as high as skyscrapers
Car driving is pretty close, but I think the combo clearing is due to RNG.
As for the super speed, the game caps how long you gain extra speed from skitching, it's just that when you grind after skitching it adds on additional speed for whatever reason.
For the street warriors goal, I tried my best to imitate what you did in your video. But, I just couldn't get a couple of the bounces to land right after several hours of trying, so I just settled on a low-ish :46 time. It could definitely be better, yeah, but at that point I already scrapped and restarted like five times so I just wanted to continue.
Call-out tricks are RNG dependent, but there are a number of factors that play into it:
1. Your initial set is dependent on an RNG seed that changes every ten or so frames. I played around with a number of seeds, but I couldn't get a good seed that didn't waste time for waiting.
2. Later sets actually change, surprisingly. For example, if you do all of these tricks and land on frame 1000, then you might get a Melon. However, if you optimize it and get as 993 or something, it might be a Kickflip instead. So, it is extremely difficult/time consuming to figure out what tricks to do at what times, when to land them, and stuff like that. A more optimized NJ TAS would definitely route that out better.
I figured out later that I could've buttslapped up the ramp and cleared them a lot faster, but it was too far down the line for me to go back. But, again, this goes back to me saying screw it and wanting to continue the TAS (since I already spent three days in NJ alone).
The speed thing has been known for a good while now, and we actually do them in RTA if the car RNG is in our favor! But, as Fog mentioned, there is a sort of time limit to speed that you currently can't manipulate. It would be cool if we could keep it, but nothing known increases the timer.
Also, glad that everyone is enjoying it! :)
If all goes well, it should be up in a few hours or so. :)
Samsara wrote:
Hype run, absolutely voting Yes. There's a tiny problem in that the file ends before the credits: If you provide a new file that does reach the credits, I can replace it.