Posts for xenos

Experienced Forum User, Published Author, Player (119)
Joined: 1/30/2007
Posts: 82
Comicalflop wrote:
And Green shells won't be any faster, since they have to bounce off of something first and "fire and it homes back at you" will be far easier, and faster, than "bounce off a wall at the precise angle to hit you".
I should probably point out that the red shells don't actually home back on me but on one of the CPU's and I just happen to get in the way so there is equal precision there. If green shells were to able to get you over the wall they would be faster because you can collect three of them at once saving you the second item box trip. So anyway, I discovered a small technique to speed you up while in a slide; all you need is to move the control stick back and forth every frame. This technique has limited use because you would typically be charging an MT instead. It is useful while shrooming, though. To prove its usefulness I attempted to beat Fried's MR flap and just barely succeeded. Video. Also, I haven't seen it mentioned yet so here is the speed address for memory watching - 00A5121D. Yay! No more speedometer.
Experienced Forum User, Published Author, Player (119)
Joined: 1/30/2007
Posts: 82
I found a somewhat new strategy on LR that is faster than the current one though I wouldn't be surprised if someone had already thought of this given how obvious it is. Using red shells from the tunnel, you can launch yourself over the wall in much the same way that blue shells do. The end result is, as Weatherton predicted, much less entertaining than the current run which is a bit of a shame given how much I enjoyed the previous. I made a test run that finishes in 35.31.
Experienced Forum User, Published Author, Player (119)
Joined: 1/30/2007
Posts: 82
Comicalflop wrote:
Ok, new topic. Way back in February I was working on a new method for doing the 1st lap in Luigi Raceway. Here it is. I can't quite seem to find a way to build up enough speed to mylestyle over the sandy hill; however, if I held AB and kept speed brake spinning, I may be able to maintain more momentum after landing and build enough speed to get over the gray wall.
You don't need that much speed to do the mylestyle. I made a short movie that completes the first lap in 6.76 which can be found here. Time is wasted at the start because I was testing the idea of doing the lap 3 trick on lap 1 (which was very difficult). Edit: Although it doesn't really matter because you have to wait for Wario on lap 3 anyway.
Experienced Forum User, Published Author, Player (119)
Joined: 1/30/2007
Posts: 82
I thought I'd give this game a whirl just for fun. I time trialed Choco Mountain which, admittedly, is probably not as exciting as GP but I still enjoyed it in any case. Link.
Experienced Forum User, Published Author, Player (119)
Joined: 1/30/2007
Posts: 82
ShadowOfMyles wrote:
The jetpack is awesome! How exactly is it executed though? I mean how does dash turn sideways and fly through the wall like that?
It's super easy to do. There is a description of the trick here from its discoverer although for one reason or another he didn't realise the trick could be performed anywhere. The mechanism behind being able to move through the walls is no doubt similar to that of SM64's BLJ however, some walls seem to be "immune" to speed.
Experienced Forum User, Published Author, Player (119)
Joined: 1/30/2007
Posts: 82
I made a TAS of the sewers of the imperial city this time. This is probably one of the more interesting levels and I managed to find a few time savers along the way. http://www.youtube.com/watch?v=CT2DKbrDqqs
Experienced Forum User, Published Author, Player (119)
Joined: 1/30/2007
Posts: 82
It turns out that the wall climbing glitch can be done anywhere and thanks to the geometry of Gall Spaceport, you can skip a great deal of the longest level in the game. Youtube link.
Experienced Forum User, Published Author, Player (119)
Joined: 1/30/2007
Posts: 82
I don't think this game has good enough emulation to do a full run. Desynchs are a problem and when I play the game I often get error messages from napalm about the textures. You may also notice that the explosion textures do not emulate correctly but are instead replaced with other textures. It would probably be worth waiting for the new version of mupen to see if that fixes the problems.
RT-55J wrote:
Do you have any estimate on how long a full run would be?
My guess is around 45 minutes, though it has been a while since I played this game.
Post subject: Star Wars: Shadows of the Empire
Experienced Forum User, Published Author, Player (119)
Joined: 1/30/2007
Posts: 82
Tool assisted speedruns: Battle of Hoth (Jedi) 2:19/m64 Echo Base 3:01 Gall Spaceport Part 2 1:53 (approx) Sewers of the Imperial Palace 2:18 List of real time speedruns: Echo Base 3:24 The Asteroid Field 3:14 Mos Eisley and Beggar's Canyon 2:38 Imperial freighter suprosa 2:49 Sewers of the Imperial City 4:42 Xizor's Palace 4:35 Old Text: I made a test run of Echo Base in "Shadows of the Empire" to determine whether this game is suitable for TASing. This game has a reasonable amount of variety in the levels so I think it is. Obviously there are emulation problems at present so desynchs are a given but the advent of Napalm has at least seen a fix to a major graphical error (or fixed from whatever plugins I was using before).
Experienced Forum User, Published Author, Player (119)
Joined: 1/30/2007
Posts: 82
I made a TAS one of my favourite courses - Red Canyon. I found no knew strategies but it's still fairly quick. Link
Experienced Forum User, Published Author, Player (119)
Joined: 1/30/2007
Posts: 82
A problem I came across a while ago. Calculate the average distance between two points on the surface of the unit sphere. That is, if two points were placed on the surface of a sphere with a radius of one and a chord drawn between them, the average length of the chord if the points were selected at random.
Post subject: Re: Improving the time
Experienced Forum User, Published Author, Player (119)
Joined: 1/30/2007
Posts: 82
Dagni wrote:
I'm 98% sure that you can shave 1.5 or 2 full seconds off the time, though. It's been a long time since I played deathrace (best is about a 45 second time, not tool-assisted), but I'm almost completely sure that if you fall off the course and accelerate to your death ASAP the clock stops as soon as you die, saving a little time because the clock is no longer running while waiting for all the enemies to finish dying and the game to declare you a winner. Just by quickly watching the vid and pausing, it looks like at about 6.8 seconds you have all 29 stars and are very near an open edge.
I realised this just after I posted the video so I edited the movie description. I modified the ending to get a time of 6"993. Time could still be saved elsewhere. If I could manipulate the drivers better I wouldn't have needed to slow down at the end to kill the drivers behind me. The time could probably go under 6 seconds this way.
Experienced Forum User, Published Author, Player (119)
Joined: 1/30/2007
Posts: 82
Progress report. I redid the second half of the boss and then did JF and AL. The next update should be at the start of SM. The M64 and the state on FF.
Post subject: [Post missing]
Experienced Forum User, Published Author, Player (119)
Joined: 1/30/2007
Posts: 82
[Post content lost in a database crash]
Experienced Forum User, Published Author, Player (119)
Joined: 1/30/2007
Posts: 82
I redid the deathrace 'cause the other one was just too slow. Youtube. If anyone wants the .m64 of the run I could upload that too.
Experienced Forum User, Published Author, Player (119)
Joined: 1/30/2007
Posts: 82
Just to clarify, the current run is aiming to defeat Wizpig I as soon as possible to reach the first credits. This means no trophy races and no TT stages (Icicle Pyramid etc). Predicted time at this point being just under 1:20 my guess being 1:17.
Experienced Forum User, Published Author, Player (119)
Joined: 1/30/2007
Posts: 82
Just a small update. I finished HTV 9 seconds faster than the previous. I always thought JF would be the best candidate for a double lapping but evidently it is not so. The boss was also done but I will likely redo the second half so I can at least gain 1 second over the previous (it is just over 0.5 at the moment). Thus far 50 seconds have been gained compared to version one up to the same point though they can not be directly compared in this way. M64 of the run.
Experienced Forum User, Published Author, Player (119)
Joined: 1/30/2007
Posts: 82
Coincidently I was recently playing Super Mario Land at higher speeds to improve my skill. I did level 1-1 at 200%, the others are too hard at this speed. Here is a video of that first level.
Experienced Forum User, Published Author, Player (119)
Joined: 1/30/2007
Posts: 82
comicalflop wrote:
JKT and I discussed about the 360 degree spin strategy, which would theoretically achieve 5-6 seconds. Want to give it a try Xenos?
Well, I tried the spin attack but I couldn't make it work. I wasn't able to move from side to side quick enough to catch all the racers. Also, perhaps JKT should try this given that he knows significantly more about this game than I do.
Experienced Forum User, Published Author, Player (119)
Joined: 1/30/2007
Posts: 82
Derakon wrote:
Sorry, I wasn't asking the right question. Why was it necessary to use a spin attack at the end to catch someone behind you? Shouldn't you be killing every car as you pass it? Mind you, I don't know that it's worth a slightly faster time if you have to give up the announcer's lines. :)
The person up ahead was moving too fast. Had I have gone to the final racer first I would have lost a lot of ground over the racer up ahead and as you can see I only just caught up to the front racer moving at full speed off the jumper. Also, I would have needed to cross the finish line.
Experienced Forum User, Published Author, Player (119)
Joined: 1/30/2007
Posts: 82
After my second attempt at this I clocked a 13"980. I tried the tactic of going in front of everyone and spinning but it wasn't very effective or I was not doing it correctly. It turned out pretty well with the current strat anyway.
Experienced Forum User, Published Author, Player (119)
Joined: 1/30/2007
Posts: 82
Bisqwit wrote:
God created the world, and God can see the future, but neither of them robs the human of free will.
True. Knowledge doesn't negate free will but the two ideas put together might. If it is true that God is both omniscient and omnipotent and created the universe this could lead to a conclusion of no free will. Consider this; before the universe was created, God knew he would create a universe and would know everything that would happen inside that universe. Looking at this from a different perspective, it is as if God created the universe with a specific design in mind. In this way, everything is playing out exactly the way God wanted things to. This would mean that free will is non existent because all actions God foreplanned. It is like if I created a robot designed to kill. Who is responsible for the actions of the robot? The robot or the creator. I created the robot knowing exactly what it would do. I must therefore be responsible. How is this situation different from the former? More on topic: What do you think would happen if two omniscient beings played each other in a game of chicken?
Experienced Forum User, Published Author, Player (119)
Joined: 1/30/2007
Posts: 82
Hina98 wrote:
Does the trick have anything to do with the honking? It looks like you get a little bit of an extra boost. It seemed like you took the angle a little wide on the second or third lap when coming up to the pair of missile and boost balloons. Was this faster?
The trick was collecting both blue baloons at once. Changing the speedometer seems to make me go slightly faster in some places some of the time but the map is more interesting than the speedometer anyhow.
Soulrivers wrote:
I must however point out that doing a 100% run is incredibly boring. Why not aim for beating Wizpig as fast as possible? I think that'd be much more entertaining than watching all levels three times over.
Indeed. The new goal is to now get to the first credits as fast as possible making it more of an any%. The new route should take about 1:20 unless someone really thinks the old objective was better. I redid the final lap of AL to reflect this route change and completed FC nearly 13 seconds faster than the previous version. M64/Savestate
Experienced Forum User, Published Author, Player (119)
Joined: 1/30/2007
Posts: 82
Ancient Lake done. 8.4 seconds faster in total. Only one track more but it took me a long time and it kills the old version due to a new trick so I thought I'd post it.
Halamantariel wrote:
My only concern is the end of the 2nd lap, at the "bump". You get a smaller bump but you get more bananas. IIRC, after 10 bananas, there is no more speed gain. Therefore, it might have been a better choice to get a bigger bump (thus keeping your speed for a longer time) instead.
I couldn't make the jump any larger because of the CPU's creating lag which makes the jump smaller for some reason.
Hina98 wrote:
It looked much cleaner than your first version did. Was the 3.8 seconds primarily from gathering the golden balloons outside of the courses or was it from the first track?
2.5 seconds from the baloon collecting, 1.5 from JF.
comicalflop wrote:
I'd recommend doing the same thing and submitting a faster, better, more entertaining version 2; it will have a much better reception. Getting memory addresses shouldn't be too hard with L-Spiro's MHS (I'd recommend speed, x/y/z pixel positions, etc.)
I tried but it did not work too well. I don't think the memory adresses will be of great help.
Experienced Forum User, Published Author, Player (119)
Joined: 1/30/2007
Posts: 82
Does this look any better than the previous run even though it is a meagre 3.8 seconds better? That's not to say I will redo the run I was merely bored one day.